PK3 2nded ReferenceManual v1.1 Virtual

PRE-K – 3RD
REFERENCE
MANUAL
CLASSROOM ASSESSMENT
®
SCORING SYSTEM
2nd Edition
®
PRE-K–3RD
Reference Manual
Classroom Assessment
Scoring System®
2nd Edition
®
®
Copyright © 2022 by Teachstone, Inc. All rights reserved.

All images © Teachstone, Inc. or with permission from licensed stock image libraries.
“Teachstone,” “Classroom Assessment Scoring System,” and “CLASS” are trademarks of Teachstone, Inc.
No portion of this document may be reproduced or distributed for any purpose without prior written permission
from Teachstone, Inc.
Contents
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1
Preface: CLASS 2nd Edition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
Chapter 1: CLASS—An Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
What is the CLASS CQI System? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
What does CLASS measure? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
What CLASS Does Not Measure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
Theoretical and Empirical Foundations of CLASS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
How was CLASS developed? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
Chapter 2: CLASS Research Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
CLASS and Child Outcomes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
CLASS across a Diversity of Populations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Improving the Quality of Interactions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
Research Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
Chapter 3: CLASS in Policy and Practice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
CLASS and the Office of Head Start . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
CLASS and QRIS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
CLASS in Pre-K–12 Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
Chapter 4: Establishing a CLASS Observation and Reporting System . . . . . . . . . . . . . . . . . . . 31
CLASS Observation Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Supporting Readiness in the Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Developing an Observation Protocol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
Observation Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
Maintaining a Pool of CLASS Observers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Data Quality Checks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
Providing Feedback to Educators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
Chapter 5: Promoting Equitable Uses of CLASS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Equitable Observation Practices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Appendix: CLASS Technical Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
General Properties of the CLASS Measure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
Reliability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
Validity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
®
Acknowledgments
The introduction of CLASS in the early 2000s focused the attention of researchers, practitioners, and
policymakers on the value of interactions between educators and children and helped elevate the importance
of classroom practices more broadly. CLASS 2nd Edition builds on this contribution by refining the specific
definitions of interactions. This edition also better represents the ways in which the measurement and
improvement of interactions must reflect the diversity of settings in which young children spend time.
CLASS 2nd Edition is a team effort, no longer reflecting the work of a small set of authors. Rather, CLASS
2nd Edition is a product of collaborations with educators, researchers, and policymakers throughout
the world. As such, this edition does not designate specific authors. The publisher is Teachstone—the
certified B-Corporation we founded to support the use of CLASS to improve the quality and impact of
early education opportunities. However, it is important to recognize the many people who contributed
their time, talent, and ideas to the effort. Because we cannot name everyone (there are too many), we
use this opportunity to thank the communities that contributed to this project to improve CLASS and its
contributions to the field. This includes scholars, educators, our network of advisers and consultants, and
our colleagues at Teachstone.
Two threads run through this work: the life-changing power of relationships and the collective impact
achieved when a broad and diverse community, sharing a high-level vision around supporting children to
reach their full potential, challenges and learns from each other. The relationships we’ve formed, and the
communities in which we do our work, are the true authors of this second edition.
Scholars
The earliest work on CLASS was primarily conducted by a small group of researchers involved in the
National Center for Early Development and Learning (NCEDL) Study of State Pre-K. We built on the work
one of us (Dr. Bob Pianta) led with the National Institute of Child Health and Development (NICHD) Study
of Early Child Care and Youth Development. These researchers from the University of Virginia, University
of North Carolina, and University of Los Angeles ushered in a fundamental shift in the definition of quality
in early childhood education. Their work focused on the most essential elements of children’s experiences
of interactions with adults and peers. Findings from the NCEDL, alongside those from the NICHD,
demonstrated that variation in features of educator-child interactions—particularly emotionally supportive,
well-regulated, and cognitively stimulating exchanges—resulted in positive shifts in children’s development
and learning across social and academic domains.
After the release of these findings, many in the research community became interested in using CLASS
as a part of their own research. Starting as early as 2004, with the training of a team from the Head Start
FACES study, we began training researchers to use CLASS. In response to researchers’ requests, we
developed observation tools for additional age levels. These tools helped researchers to better understand
the contributions interactions make to learning and development in infant and toddler settings and in the
elementary and secondary school years. The hundreds of peer-reviewed studies, thousands of research
presentations, book chapters, and dissertations that have used CLASS directly inform this new edition.
Through dialogues and debates with colleagues, continued learning, and deep introspection, we expanded
our understanding of children’s classroom experiences in the United States, in other countries and cultures,
and across a broad range of academic disciplines.
Copyright © 2022 by Teachstone, Inc. All rights reserved. Pre-K–3rd Reference Manual 1
Research using CLASS has been conducted in hundreds of universities around the world. However, we owe
particular gratitude to our colleagues at the School of Education and Human Development (SEHD) at the
University of Virginia (UVA), many of whom have made foundational contributions to this effort. Scholars at
the Center for Advanced Study of Teaching and Learning (CASTL) at UVA conducted foundational research
on CLASS and developed and tested CLASS-based professional development across ages and grade
levels. They supported use of CLASS in international settings from Latin America to China to Europe and
worked closely with educators and policymakers who adopted CLASS as a quality improvement tool in
Virginia and beyond. Faculty in the SEHD educator preparation program supported refinements of CLASS
at the secondary level and have pioneered the use of CLASS as a part of preservice training. Economists
at the SEHD EdPolicyWorks Center helped build knowledge of CLASS as an accountability and quality
improvement tool in places such as Louisiana and Virginia. Graduate students and postdoctoral fellows
contributed to our learning as they established their own lines of research that are now adding to the
knowledge base on the power of educator-child interactions. UVA’s SEHD will always be the academic
home of CLASS and it is an honor to be a part of such an incredible academic family.
Educators and Policymakers

Although CLASS started as a research tool, it was quickly adopted by policymakers and practitioners
who saw the value of CLASS in their efforts to ensure that children had access to effective educator-child
interactions. CLASS 2nd Edition reflects the insights we gained as we have worked side by side with an
inspiring group of teachers, coaches, leaders, and policymakers.
As CLASS helped to monitor and improve quality across Head Start and many state early childhood
systems, we learned many lessons about how to ensure CLASS data is most impactful. First and foremost,
we must ensure that CLASS data provides timely, relevant, and actionable feedback to the educators
working hard each day to support young children. We look forward to continued collaborations with our
colleagues at the Office of Head Start and in states and localities across the country. These collaborations
will serve to innovate on policies and practices that help ensure educators get this feedback in the most
supportive ways for their professional learning and development.
CLASS 2nd Edition celebrates the dedication, talent, and wisdom of educators across the globe. Educators
have been our greatest teachers as we undertook CLASS revisions. They inspired us with their commitment
to the relationships they develop with children and their skills in making the most of classroom moments
to enhance children’s learning and development. It is truly an honor to elevate and give recognition to the
impact these educators have—in our personal and professional lives.
But we also know we can do more to support educators. Over the past decade, our discussions with
educators indicated that although much of CLASS resonated with them, the manuals and training materials
did not sufficiently represent many of their lived and professional experiences. The second edition
embraced the goal of having more educators see themselves in CLASS. We hope we have made progress
toward that goal.
Teachstone
CLASS 2nd Edition reflects the dedication and talent of our Teachstone colleagues. Every part of the
company influenced the changes made to the manual. At the heart of that work is Teachstone’s Impact
Department, home to our Research and Evaluation and Social Impact teams. The Impact teams did the
heavy lifting by gathering feedback from the field and spending countless hours observing, scoring, and
revising. We owe particular gratitude to Veronica Fernandez, Grace Funk, Sarah Hadden, Vicki Kintner-
Duffy, Claudia Perez, Dorothy Sanchez, and Sara Schodt for their work on the 2nd Edition, and to Sarah
Caverly for leading work on the technical aspects of the revision. Teachstone’s Content, Product, Program,
and Professional Services teams played a role in developing the manual and making sure it “works” for
observers, trainers, and educators across the world. As product manager for the Measurement Suite, Katie
2 CLASS® 2nd Edition Copyright © 2022 by Teachstone, Inc. All rights reserved.
Twilley coordinated the work and ensured alignment across all components. Marielle Sheridan led the
publishing work on this edition, and we are always appreciative of her efforts to ensure the words on these
pages are most impactful. Finally, thank you to our field-facing teams for consistently sharing the voice of
clients and partners, ensuring the work is authentic and meaningful.
Contributors, Consultants, and CLASS Advisory Board

Contributors. The community that shared in this edition of CLASS is too large to name. However, we owe
two colleagues gratitude for the specific roles they played. Dr. Stephanie M. Curenton of Boston University
and Dr. Iheoma U. Iruka of University of North Carolina at Chapel Hill supported our Teachstone team in
their early efforts to revise CLASS. They expanded core assumptions and clarified the limitations of CLASS
in measuring every child’s experience in their learning setting. We are grateful for their knowledge, wisdom,
and guidance and look forward to our continued collaborations. We know more progress needs to be made
to ensure that every child, regardless of race, ethnicity, language, gender, or disability status, experiences
the interactions they need to thrive in learning settings and beyond.
Consultants. We consulted with many scholars and educational leaders for this revision, bringing expertise
in many areas, including culturally responsive teaching across settings, trauma-informed care, social and
emotional learning, executive functioning, literacy, science, and math. We also consulted with experts in a
variety of curricular models, including High Scope, Montessori, Reggio, Tools of the Mind, and Waldorf, and
settings such as family childcare, public school, and Head Start. Each of these experts have used CLASS
and many have also developed observational tools of their own. We are grateful for their insights and
their challenges to our thinking, which opened us up to new ways of using CLASS to enhance children’s
learning and development. We reflected some of those ideas in this new edition, but others were beyond
the scope of this set of changes. We know that no single measure will ever capture everything that matters
in children’s classroom experiences, and we are heartened and inspired by the work of colleagues to inform
the next innovations in classroom measurement.
These colleagues include Barbara Acton, Rosemarie Allen, Sarah Arnold, Craig S. Bailey, Tia Navelene
Barnes, Jessica V. Barnes-Najor, Rebekah Berlin, Elena Bodrova, Alaina E. Boyle, Sonia Q. Cabell, Ann
Cameron, Christina Cipriano, Cathrine Aasen Floyd, Annie Frazer, Chrisanne Gayl, Daryl Greenfield, Mary
Louise Hemmeter, Bryant Jensen, Shannon D. Lockhart, Rafiqah B. Mustafaa, Irena Nayfield, Jelena
Obradović, Carola Oliva-Olson, Marina Rodriguez, Michelle Sarche, Wendy Shenk-Evans, Teresa Smith,
Amber Valentine-Minion, James Venza, Mavany Calac Verdugo, and Almut K. Zieher.
CLASS Advisory Board. We are also grateful to the CLASS Advisory Board. Beginning in 2020, these
partners met quarterly with Teachstone, providing guidance and feedback that was critical to the evolution
of the manual and the systems that used CLASS. Board members reflected a diverse set of perspectives
and ideas about CLASS. We are fortunate that such a distinguished group of leaders participated in this
opportunity to support systemic changes in early childhood and pre-K–12 settings by improving CLASS.
The Advisory Board includes David Adams, Laura Bailet, Ann Cameron, Christina Cipriano, Stephanie M.
Curenton, Peter Fishman, Ana Elisa Franco-Labarga, Akimi Gibson, Rolf Grafwallner, Daryl Greenfield, Gabe
Hakim, Mary Louise Hemmeter, Blair Hyatt, Iheoma U. Iruka, Cynthia Jackson, Victoria Jones, Tiffany Lee,
Lisa Luceno, Dominique McCain, Jack McCarthy, Calvin Moore, Debra Pacchiano, Erica Phillips, Paula Polito,
Natalie Renew, Tina Routh, Celia Hartman Sims, and Yasmina Vinci.
The Path Ahead

Innovation never stops. We know that as individuals and as a company, we have more work to do to elevate
the work of educators in supporting every child through their daily interactions. The promise of this new
edition of CLASS and the aligned measurement suite is exciting. These tools will help more children and
educators get the support they need to thrive. We also know that one tool can only do so much. CLASS
measurement and improvement tools are only part of a broader set of efforts to transform learning settings
so that every child’s experience is one of respect in which they are supported to relate with others,
communicate their ideas and inspiration, and think deeply about the world around them.
CLASS measurement and improvement tools, on their own, will not solve structural and systemic inequities
in our educational system, exacerbated by local, state, national, and global events in the past few years. Yet
they can be part of the solutions that help address these inequities; these revisions are a step along that
path. Alongside the research, practice, policy, and Teachstone communities, we will continue working to
identify, test, and refine resources and tools to achieve more equitable access to effective interactions and
environments in learning settings.
This 2nd Edition is the next stage of the CLASS journey. It is exciting to partner with you and so many others
as that journey continues, and we build on our relationships and knowledge to make a positive difference in
the lives of children and those who serve them.
—Robert C. Pianta & Bridget K. Hamre
Preface:
®
CLASS 2nd Edition
The Classroom Assessment Scoring System® (CLASS®) is a measurement and improvement system focused
on observing the elements of children’s learning experiences that research demonstrates matter the
most—the daily interactions among educators and children. Over the past decade, CLASS has redefined
what “quality” means in early childhood and K–12 education. “Quality” as applied to the learning setting
was historically defined by elements such as educator qualifications, ratios, and aspects of the physical
environments. Research has demonstrated that although these elements contribute to quality, they are far
from sufficient to ensure that programs promote positive outcomes for children. CLASS offers a method
for observing, measuring, and improving the educator-child interactions that directly affect children’s
development and learning on a broad scale. CLASS has helped thousands of educators ensure they are
effectively supporting the children they work with each day.
Teachstone, a certified B-Corporation founded in 2008, helps deliver on the promise of CLASS by
partnering with the field to offer the training, tools, and support needed to use the CLASS tool in the most
impactful ways. The Teachstone mission is “Helping every child reach their full potential by measuring
and improving the interactions that matter most.” Teachstone is looking to the future with an even bolder
and broader vision: “Ensuring everyone has access to life-changing interactions.” Together with the field,
Teachstone has made important progress, but there is much more to do.
The Teachstone mission is “Helping every child reach their full potential
by measuring and improving the interactions that matter most.”
Educators value CLASS as a common language and lens for describing, observing, measuring, and improving
their interactions with children. Leaders value CLASS as an anchor for policy and investments to improve the
daily experiences of children in learning settings and the effectiveness of education systems. The CLASS
system has helped improve the elements of quality that are foundational to children’s development and
learning on a very large scale in hundreds of thousands of learning settings across the globe.
CLASS has its foundations in research and theory, but its impact has come through partnerships with the
field. Those partnerships have been invaluable in informing the 2nd Edition improvements. Teachstone
colleagues in federal and state education agencies, Head Start programs, local education agencies, and
tens of thousands of educators have all advised how to best support CLASS users. They have also indicated
areas for improvement in core tools, training, and offerings. The Teachstone team listened carefully to
educators, coaches, leaders, policymakers, and researchers from across the country to find out what works
and does not work in their use of CLASS.
The second edition of the CLASS observation measure, and the aligned updates to related trainings and
tools, are the culmination of learnings from the field. Many educators around the world have invested in
CLASS infrastructure since the introduction of the measure 15 years ago. The enhancements in this 2nd
Edition make it more reflective of the field while maintaining its core structure. Teachstone will provide
technical and implementation support as the field adapts to the innovations contained in this edition.
CLASS 2nd Edition Improvements
The first CLASS Manual was published in 2008, and it is the source document used by the field to observe
and measure educator-child interactions. CLASS 2nd Edition retains the core structure and features of the
system outlined in that manual. The second edition continues to capture the components of classroom
quality that research has demonstrated matter and that minimize disruption to the field.
What Has Changed

This CLASS 2nd Edition and aligned measurement suite have three key areas of improvement:
● Centering Equity: This involves supporting observers in understanding the variety of ways that
interactions may look and sound across individuals, cultural contexts, and settings. Centering equity
also builds awareness of the ways observer bias may affect CLASS ratings.
● Increasing Access: The Pre-K–3rd tool combines the prior Pre-K and K–3 tools to reduce user burden
and cost and to support greater continuity in the measurement of quality across these critical early
years. Additionally, the information available in this Reference Manual is freely accessible as a digital
download.
● Measuring for Impact: CLASS 2nd Edition provides greater support for all users of CLASS, from
observers and coaches to leaders and educators, in using CLASS as a tool for continuous quality
improvement. As one important example, CLASS 2nd Edition now supports more systematic
assignment of ranges at the indicator level, providing more detailed and actionable feedback to
educators. Additionally, CLASS 2nd Edition includes explicit measures of children’s exposure to
content (literacy, math, social and emotional learning, science) and format of instruction (whole group,
small group, centers) as well as the CLASS Environment™ supplement.
Figure P.1 summarizes these improvements, and the following section outlines the improvements in
more detail.
Center on Equity Increase Access Measure for Impact

● More equitable and inclusive ● Easier to access ● Expanded measurement
definitions of effectiveness
о Age level simplification: о Activity Setting items
● Increased representation combined Pre-K–3rd
о CLASS Environment
● Considerations for use across о Publicly accessible Pre-K–K (optional
a diversity of contexts Reference Manual supplement)
● Support for bias recognition ● Easier to implement ● Continuous Quality
and reduction in observations Improvement
о Easy-to-use Observation
enhancements
Field Guide designed
to support the CLASS о Assigning indicator
coding process ranges to provide more
specific feedback to
о Updated score sheets
educators
designed to support the
CLASS coding process
Figure P.1. CLASS 2nd Edition Enhancements
What Has Not Changed
CLASS 2nd Edition continues to use the nested structure of domains, dimensions, indicators, and behavioral
markers to capture interactions in the learning setting. (See the Observation Field Guide for more
information on this structure and how to use it in scoring.)
In this edition, the domains, dimensions, and indicators measured during CLASS observations remain the
same. Hundreds of studies have documented that the interactions described by CLASS predict positive
learning and developmental outcomes for children across demographic and economic backgrounds.
Moreover, the structure of the tool has provided a way of describing interactions that helps to create a
lens through which to view teaching practices. Continuity in the foundational organization of the CLASS
framework allows for seamless integration of new supportive materials into existing CLASS measurement
and improvement systems. Teachstone renamed a number of indicators and dimensions to more clearly
communicate what they are intended to capture and in some cases to better reflect a more inclusive
vision of effective interactions. But these terminology improvements do not change the core elements of
observations in concept, number, sequencing, or grouping. Behavioral markers—the unscored items that
represent examples of observable interactions—contain more substantive enhancements. The Additional
Improvements section provides the details of these enhancements.
Key Improvement Details

Equity Enhancements for More Accurate Reflection of the Field. Research demonstrates the power of
interactions to enhance learning and development across contexts and cultural communities. (Refer to
Chapter 2 for a summary of this research.) However, because variations often occur in how interactions
look and sound across settings and contexts, observers must understand these variations and the ways
in which they can influence CLASS scoring. The CLASS 2nd Edition contains several important revisions
and additions:
● More equitable and inclusive definitions of effectiveness. To effectively capture the quality of
interactions that children experience, it’s critical to understand what typical interactions look and
sound like. The first edition of CLASS presented a limited vision for quality by using examples that
were typically observed in school-based settings and led by White educators. A limited range
was also present in the examples of children’s interactions and engagements. CLASS 2nd Edition
expands the definitions, behavioral markers, and descriptions of effective interactions and is more
inclusive of the diversity in educational settings.
● Increased representation. Within the observation and supplemental trainings, CLASS dimensions
have more updated examples. These examples, developed in collaboration with a diverse group of
educators with deep experience in teaching young children, more accurately reflect the range of
learning contexts that exist across cultural communities and settings.
● Considerations for the equitable use of CLASS across a diversity of settings. CLASS 2nd Edition
provides support for using CLASS across contexts by incorporating considerations that focus on
variations due to individuals’ cultural communities, language, development, or childcare setting.
These considerations help to contextualize possible variations in effective interactions that observers
may see across individuals, cultural communities, and settings and are especially important when
observers are less familiar with such contexts.
● Supporting bias recognition and reduction in CLASS observations. To fully use the second
edition improvements, observers must learn about the role that bias plays in how people view
and interact with the world around them. All people, and therefore all CLASS observers, develop
their own beliefs, experiences, and expectations about behaviors and interactions. If unaware of
these biases, an observer’s lens can bias what they notice, what they miss, their judgments about
the effectiveness of interactions, and how they assign CLASS scores. The Reference Manual
and Observation Field Guide include information about various ways that observers’ individual
experiences and beliefs can influence CLASS scores and how to reduce the impact of bias.
CLASS Observation Training, and supplemental CLASS Observation Supports trainings, will help
observers acknowledge the role that bias plays in their own thinking and behavior, learn strategies
for reducing bias, and apply those strategies in CLASS observations. No single reading or training
can fully address observer bias, and Teachstone commits to continuing collaboration with the field
to provide ongoing training and support in this area.
The improvements made to CLASS 2nd Edition support a more equitable use of CLASS to conduct
classroom observations. Additional innovation and research on CLASS and other measures of quality will
advance the field further toward the broad goals of educational equity. Teachstone commits to partnering
with the field and contributing to this critical work.
Increasing Access. CLASS is most valuable when it is easy to access and use. CLASS 2nd Edition includes
revisions that increase ease of access and implementation for educators, programs, and systems to benefit
children, families, and communities.
● Age Level Simplification. Many measures of quality—particularly those used in K–12 settings—rarely
acknowledge adjustments to teaching practice that correspond to children’s development over time.
CLASS has always emphasized developmental variation in interactions to align to the science on
human development and provide the most relevant feedback to educators. However, multiple age-
related versions place burden on observers who need to train and certify across age levels and on
systems that support observers on multiple levels. This Pre-K–3rd tool combines the prior Pre-K and
K–3 tools to reduce burden and cost and to support greater continuity in the measurement of quality
across these critical early years.
● Publicly Accessible Reference Manual. To help organizations and programs use the CLASS
system most effectively, Teachstone created the CLASS 2nd Edition Reference Manual to support
understanding of CLASS and its implementation. This information is available in print and as a free
download to ensure all programs can access it.
Measuring for Impact. Capturing CLASS data makes a difference when used to support educators at all
levels to enhance their practice. CLASS 2nd Edition provides important improvements that help ensure
more effective use of CLASS observations as a continuous quality improvement (CQI) tool. Additionally,
although interactions remain at the core of CLASS, children’s learning and development is also affected by
the activity settings, content, and environments in which these interactions occur.
● Include Activity Setting items. The revision of these items helps observers better record critical
information about children’s access to and engagement in content, such as literacy, math, and social
and emotional learning, and the format of content instruction, such as small group, whole group, or
centers. These data provide a more comprehensive picture of children’s experience with curricula in
the context of interactions.
● Add CLASS Environment. This new measure supplements CLASS and provides a more streamlined
approach to measuring and improving the most important elements of the environment that promote
effective educator-child interactions. CLASS Environment is designed for use alongside CLASS.
● Require indicator range assignment. CLASS 2nd Edition formalizes the assignment of indicator
ranges (low, mid, high) during scoring. Although many observers already assign indicator ranges,
some do not. The Observation Field Guide describes a more detailed process of scoring,
benchmarked to indicators, to increase the rigor and accuracy of scoring and to provide more
actionable feedback to educators.
● Include information to guide CLASS program and classroom reporting. Teachstone provides
guidance on using CLASS to support systematic improvement at the program and classroom level
and links to resources to enhance reporting (see Chapter 4).
Additional Improvements
Beyond these key changes, other notable improvements include the following:
● Revisions to the language on the face pages. Although the core concepts and framework of CLASS
remain the same, much of the language has been revised, such as rewording, removing, or adding
behavioral markers or definitions and refining indicator names. Each update falls into one or more of
the following categories:
о Language updates to more accurately reflect variation in interactions across contexts
о Language updates to provide clarification to observers and remove academic jargon
о Addition of behavioral markers to better represent the indicator
о Modifications to domain and dimension definitions to align with the improvements listed above
о Addition of indicator definitions to support coding
● Revisions to the indicator descriptions. The newly created Observation Field Guide provides
indicator descriptions that align with revisions to the indicators and behavioral markers and that are
more applicable across multiple settings. These generalized descriptions help observers decide
on the most appropriate range for each indicator during the scoring process. CLASS Observation
Training and other resources include more specific examples of the indicators.
● New guidance on effectively collecting CLASS data. Learning how to assign CLASS scores with
consistency is just one part of ensuring observers are ready to collect CLASS data. CLASS 2nd
Edition provides enhanced guidance for how to collect data across settings, using live observation or
video, as well as new tools that can enhance data collection efforts.
● Updated CLASS Technical Summary. A new technical summary, at the end of this manual,
summarizes the most up-to-date research evidence on CLASS and provides information on its
technical properties.
● Creation of an easy-to-use Observation Field Guide. The CLASS 2nd Edition Observation Field
Guide was specifically designed to enhance observer understanding and support the coding
process. Separate from the Reference Manual, this field guide explains key terms; provides details
about each dimension, indicator, and behavioral marker; and provides guidance on reducing bias
in observations and addressing common observation questions. To score appropriately, observers
must reference the field guide for every CLASS observation.
Aligned improvements to other elements of the measurement suite enhance the impact of these
improvements to the CLASS Manual. These enhancements include:
● Updated training and certification programs
● Updated score sheets
● Improvements to the CLASS Data Collection and Reporting tools on the myTeachstone® platform
● Supports for Affiliate CLASS Trainers and the CLASS Learning Community
Up-to-date information about these enhancements can be found at teachstone.com.
®
Introduction
This CLASS 2nd Edition Pre-K–3rd Reference Manual includes summaries of key
research, evaluations of large-scale implementations, guidance on creating a data
collection system, and support for reporting CLASS data. This manual does not
provide the information necessary to conduct CLASS observations; the CLASS
2nd Edition Pre-K–3rd Observation Field Guide contains the assessment tool that
Certified CLASS Observers use to conduct CLASS observations.
®
Chapter 1:
CLASS—An Overview
The Classroom Assessment Scoring System (CLASS) is an observation instrument developed to assess
quality in learning settings1 serving infants through high schoolers. CLASS is the most-researched and
validated observational system for measuring and improving educator2-child interactions and is widely used
across the early childhood field and in K–12 settings.
Across age levels, CLASS describes and supports observation, measurement,
and improvement of key features of educator-child interactions. CLASS
focuses on educators’ provision of emotionally safe and supportive
interactions; support of children’s behavior, time, and engagement; and
facilitation of children’s cognitive and language development. These features
of the learning setting support development and learning across the age
spectrum, from infancy through secondary education. However, the specific
nature of these interactions shifts across age levels. Therefore, CLASS
maintains a focus on key elements of interactions from birth through high
school settings, but offers differentiated age levels to ensure observations
are sensitive to the specific needs of older children.
Educator-child interactions are at the heart of CLASS because they
are the most essential element of classroom3 experiences. CLASS also
enables educators to measure and improve other key aspects of children’s
daily experiences in learning settings, including the activity settings and
environments in which interactions occur. Together this broad picture of
classroom life allows educators to understand the ways in which curriculum
and instruction come together to enhance children’s learning and
development (see Figure 1.1).
Before focusing on the nuances of the CLASS measure, it is important to Figure 1.1. Aspects of Classroom Quality
understand the way it fits into the broader CLASS CQI System.
What is the CLASS CQI system?

Programs and school districts across the country use CLASS to help
ensure all children in their programs have access to effective educator-
child interactions. This CLASS Reference Manual is part of a larger CLASS
CQI System. This system supports educators in focusing on, measuring,
and improving the elements of classroom experience that matter most for
supporting children’s learning and development (see Figure 1.2).
Figure 1.2. The CLASS CQI System
1
The term learning setting is used broadly to refer to the spaces in which educator-child interactions take place. This includes spaces such as
family childcare homes or outdoor learning environments in which educators are caring for and teaching children.
2
he term educators describes the primarily non-familial adults who work in classrooms or home settings with children. Many, but not all, of these
T
educators refer to themselves as teachers. The term educators is more inclusive and ensures that all educators who work closely with children
in these ways see themselves reflected in CLASS.
3
The term classroom is used to refer to the whole group of educators and children within a learning setting.
When all three elements of the system come together, with interactions at the center, they create a cycle
of CQI that is anchored in research and best practices.
CLASS Measurement Suite

The CLASS Measurement Suite is a core part of the
CLASS CQI System, designed to ensure educators can
effectively and efficiently measure classroom quality
in ways that directly inform improvement efforts. The
CLASS Measurement Suite has four key elements (see
Figure 1.3): the CLASS Manual Set, CLASS Observation
Training and Certification programs, CLASS Data and
Reporting tools, and CLASS Community and Affiliate
programs. Related products and services can be found
at teachstone.com.
CLASS Manual Set. The CLASS 2nd Edition Pre-K–3rd
Manual Set consists of two separate resources: the
Reference Manual, designed to provide foundational
information on the development and use of the CLASS
measure, and the Observation Field Guide, used to Figure 1.3. The CLASS Measurement Suite
guide observation and coding in the learning setting.
Observers will use both parts of the Manual Set in
training but will only need the field guide to observe learning settings. Educators and leaders may also use
the field guide to understand the overall framework and to improve educator-child interactions.
CLASS Training and Certification. Teachstone requires training and certification on the use of the
Observation Field Guide for all observers. The training helps observers understand and use the CLASS lens
and language to assess educator-child interactions. The online certification process ensures that observers
can use the tool in a reliable and accurate manner across settings and contexts.
CLASS Data and Reporting Tools. Data entry and reporting tools within the myTeachstone online platform
help to ensure accurate data entry and secure data storage and provide feedback reports to educators and
program coordinators. These reports can inform classroom and program CQI efforts.
CLASS Community and Affiliate Programs. These programs provide continued support to CLASS users
in the field. The CLASS Learning Community is an online space where observers can connect with other
educators, share ideas, and ask questions to build their CLASS knowledge. Teachstone certifies Affiliate
CLASS Trainers to deliver the CLASS Observation Training and a variety of professional development
products and services to observers and educators. They also act as in-program experts, helping their
colleagues deepen their CLASS knowledge and skills.
What does CLASS measure?

CLASS focuses on measuring classroom interactions, defined as the moment-to-moment, back-and-forth
exchanges (both verbal and nonverbal) that build and sustain relationships and support child development
and learning. The CLASS measure uses behavioral evidence from both educators and children in the
learning setting to gauge the effectiveness of such interactions by considering the depth, frequency,
and duration of observed behaviors. In many instances, the observer may need to look at the impact the
educator-child interaction has on the child to determine its effectiveness.
The Pre-K–3rd CLASS tool measures three broad domains of educator-child interactions: Emotional
Support, Classroom Organization, and Instructional Support. Within each of these domains are multiple
dimensions of educator-child interactions. Table 1.1 summarizes these domains and dimensions.
Table 1.1. CLASS 2nd Edition Pre-K–3rd Domains and Dimensions
Emotional Support
The domain of Emotional Support assesses educators’ support of children’s social and emotional functioning in
the learning setting. Educators understand that children are developing their sense of self and relationships with
others in their communities, and they support this development in a responsive way.
Educators foster connections and a sense of belonging between adults and children,
among peers, and as a classroom community. Verbal and nonverbal communications
Positive Climate between educators and children and among children convey warmth, respect, and
collaboration. The educators’ interactions enhance each child’s enjoyment of the learning
setting and their experience of it as a caring community.
Educators and children express little relational negativity verbally or nonverbally.

Educators and children rarely display irritability, anger, or disrespect toward others.
Negative Climate
Educators do not enact threats or severe punishment that cause disruptions to
relationships in the learning setting.
Educators are aware of and responsive to children’s needs—social, emotional, physical,

Educator academic, linguistic, and cognitive. The educators’ sensitivity supports children’s feelings
Sensitivity of safety and comfort in the learning setting and facilitates children’s ability to actively
participate, explore, and take risks.
Educators emphasize children’s emerging sense of self and help children develop and
express their unique interests, motivations, and points of view by providing opportunities
Regard for Child
for children to experience autonomy and direct their own learning. Children’s interests and
Perspectives
choices guide classroom experiences and, as a result, children are meaningful contributors
to activities.
Classroom Organization
The domain of Classroom Organization assesses educators’ support of children’s behavior, time, and engagement
in the learning setting. Educators understand that children are learning how to organize and manage their time,
attention, and behavior to achieve tasks, and they provide support for that learning.
Educators support children’s growing behavioral regulation skills by creating

developmentally informed, clear, consistent expectations and proactively supporting
Behavior
cooperative behaviors. Children may demonstrate challenging behaviors as they learn
Management
these skills, but educators’ methods for preventing and positively redirecting these
behaviors result in the occurrences being infrequent, mild, and quickly addressed.
Educators use time and structure activities, routines, and transitions so that children have
Productivity
regular, ongoing opportunities to participate and know how to do so.
Educators facilitate activities by supporting work and play in ways that enhance children’s
engagement. Educators balance this facilitation with moments of observation as children
Instructional engage in independent or peer play or work. Educators support children’s general
Learning Formats engagement and enhance their focus on specific learning objectives within activities.
Through these efforts, children remain deeply engaged in work and play, as demonstrated
by their active participation and focused attention.
Instructional Support
The domain of Instructional Support assesses educators’ support of children’s cognitive and language
development. Educators understand that each child is capable of learning, and they provide individualized
challenges, scaffolding, and feedback.
Educators use instructional strategies and activities that help children learn about and
understand concepts and content. Educators facilitate learning opportunities that support
children’s development of thinking skills and creativity. Factual information is taught in
Concept
the context of these learning opportunities rather than in rote ways that focus only on
Development
memorization or recall of information. Educators help children create meaning by linking
new concepts and content to prior knowledge and ensuring it is connected to their lived
experiences.
Educators provide feedback that builds on children’s knowledge and skills in ways that
expand understanding or increase persistence. Effective feedback is extended, specific,
Quality of
and individualized, meeting children where they are and scaffolding support as children
Feedback
deepen and refine their learning. Educators also enhance children’s motivation and
persistence by encouraging and affirming their efforts rather than their work products.
Educators promote and expand children’s language development and verbal and nonverbal
communication skills. Educators support children’s development in both the language(s) of
Language
instruction and children’s home language(s). Educators encourage conversations, provide
Modeling
individualized language support, and use varied descriptive language such that children
understand and communicate more in the learning setting.
Dimensions are then broken down into four or five indicators, or categories, that collectively illustrate
the types of interactions measured by the dimension. Indicators are further broken down into behavioral
markers, which are examples of potential behaviors that would count as evidence for each indicator.
Age Ranges. The Pre-K–3rd CLASS tool captures interactions in preschool through third-grade learning
settings, with children between the ages of three and nine. This age range represents a wide variety of
developmental abilities, and some examples or behavioral markers are more applicable to certain ages or
grades than others. However, the dimensions and indicators apply across ages and developmental stages.
The tool provides examples from various ages where relevant. Other CLASS tools are specifically for
younger (Infant and Toddler) and older (Upper Elementary and Secondary) children. In some cases the Pre-
K–3rd CLASS tool could be used in grades 4–6 if needed for a singular tool across a school. However, the
Upper Elementary CLASS tool better captures the developmental variations of children in middle childhood
and includes a more detailed focus on content instruction techniques.
Activity Setting Items. The CLASS Observation Score Sheets provide space to capture children’s exposure
to activity settings, including the amount of time, number of children, and engagement level in content areas
(such as literacy, math, and social and emotional learning) and formats (such as whole group, small group,
and centers or free play) during interactions. Table 1.2 summarizes the details collected by the Activity
Setting items. The items and rating process can be found in the Observation Field Guide.
CLASS Environment. The Pre-K–K CLASS Environment supplement measures elements of the environment
in preschool and kindergarten learning settings that support meaningful, impactful interactions as seen
during observation periods. Detailed information on the CLASS Environment can be found in the CLASS
Environment Manual.
Table 1.2. Details Captured by the Activity Setting Items
Content Time Number Engagement Format

The subject area or The relative The relative The degree of The format in
topic of classroom time spent in a number of children engagement in a which children’s
activities particular content involved in a particular content content learning
area or content- particular content area or content- occurs
related learning area or content- related learning
activity related learning activity
activity
What CLASS Does Not Measure

It is beyond the scope of any one tool to measure every element of the learning setting that is important
to children’s development and learning. Educators who wish to find measures that focus on these other
aspects of children’s classroom experiences can look to educational research literature. These include
content-specific instructional practices in areas such as literacy (Smith & Dickinson, 2002), math (Clements &
Sarama, 2014), and social and emotional learning (Denham et al., 2012). CLASS captures a holistic picture of
classroom quality but does not provide specific information about individual children’s experiences in ways
that other measures are specifically designed to assess (Downer et al., 2011).
Research demonstrates that CLASS captures teaching practices that support learning and development
across cultural communities and contexts. However, it was not designed to measure the experiences of
specific populations of children. Tools designed to focus exclusively on practices for DLLs (Oliva-Olson et
al., 2021), children with disabilities (Soukakou, 2016), or racially or ethnically minoritized children (Curenton
et al., 2019),4 help educators measure classroom experiences for these specific populations in ways that
complement more global measures such as CLASS.
Theoretical and Empirical Foundations of CLASS

Developmental and educational psychology theory and research are the basis for the dimensions and
indicators of the CLASS tool. At the most fundamental level, the basis of CLASS theory demonstrates the
value of the interactions children have with adults. That theory defines these interactions by the properties
of “serve and volley” exchanges that require attention to both the educator’s behavior and the child’s
response (Sroufe, 1996). CLASS and the aligned Teaching Through Interactions framework (Pianta & Hamre,
2009) offer an organization system and set of definitions of educator-child interactions. Different from
measures organized as lists of discrete behaviors (for example, smiling, attending to a child), CLASS helps
organize these behaviors into meaningful categories to enhance observation and feedback to educators.
Studies across age levels and cultures (Hafen et al., 2015; Hu et al., 2016; Leyva et al., 2015; NICHD ECCRN,
2003; Sandilos et al., 2017; Slot et al., 2017) validate this structure of CLASS, with dimensions of interactions
organized into three broad domains. The following section provides a high-level summary of the theoretical
and empirical basis for CLASS. Chapter 2 provides a summary of CLASS research of children’s development
by domain. More detailed reviews are available elsewhere (Hamre, 2014; Hamre & Pianta, 2007; Pianta &
Hamre 2009; Pianta et al., 2020a).
4
The Assessing Classroom Sociocultural Equity Scale (ACSES) measures equitable sociocultural interactions in early childhood settings and is
used in settings with varying racial or cultural compositions.
CLASS Domains
The ten CLASS dimensions are grouped into three broad domains that reflect children’s core needs and
skills such as children’s abilities to learn to relate, regulate, think, and communicate. Specifically, the domain
of Emotional Support addresses how educators support children to develop their sense of self and learn
to relate to others. The domain of Classroom Organization captures how educators help children learn to
regulate their behavior, time, and attention. The domain of Instructional Support assesses how educators
promote children’s thinking and communication skills. The interactions described by each domain and
dimension support key skills needed for continued success in school and beyond. The following sections
describe research highlighting aspects of these domains and their links to children’s outcomes.
Emotional Support
When educators develop positive relationships with children and are sensitive to their cues and needs,
children gain important prosocial (Johnson et al., 2013) and self-regulatory (Williford et al., 2013) skills.
Educators’ emotional support can also buffer children with behavior problems from the way those problems
hinder learning (Domínguez et al., 2011). Settings with stronger emotional support also enhance academic
learning. Children in such learning settings demonstrate enhanced reading development in the early years
of schooling, particularly among dual language learners (DLLs) and children at risk for reading difficulties
(Lopez, 2012; Silinskas et al., 2017). The consistency of emotional support is also important: children in
settings in which educators are more consistent in providing emotional support across a day gain more in
early academic skills (Curby et al., 2013). Emotional support may affect children, in part, through the stress
response system. Stress typically increases throughout the day for children in out-of-home settings, but not
when they are in learning settings in which educators offer warm, supportive care (Hatfield et al., 2013).
Classroom Organization
Organizational and managerial aspects of the learning setting help young children develop self-regulatory and
executive functioning skills and help set the stage for engagement in learning and curricular activities. Young
children are better able to control their behavior and cognition in the setting, and spend less time off-task,
when educators manage behavior and attention proactively (Rimm-Kaufman et al., 2009). The value of a well-
organized and managed learning setting apparently extends beyond direct effects on children’s self-regulatory
skills. Children in these types of settings also have stronger early cognitive and academic development
(Downer et al., 2012). In one study among Head Start programs, children demonstrated stronger early writing
skills, phonological awareness, listening comprehension, alphabet knowledge, and book knowledge, when
levels of Classroom Organization were higher (Maier et al., 2012).
Instructional Support
Instructional support are the strategies educators use to promote children’s cognitive and language
development. Educators can facilitate higher-order thinking skills and knowledge, provide feedback, and
scaffold vocabulary and communication skills. Children exposed to these rich instructional interactions
demonstrate stronger development across literacy, math, science, and social and emotional domains
(Hamre et al., 2014; Howes et al., 2008). Instructionally rich interactions can occur throughout the day, not
just during work in whole and small groups (Cabell et al., 2013). Several studies suggest that instructional
interactions may be particularly important in the development of literacy skills for children from lower-
income families (Han et al., 2017) and among DLLs (Hindman & Wasik, 2015).
CLASS Activity Settings

Although effective educator-child interactions are essential for learning and development, they are not
sufficient. Research demonstrates that for children to learn key skills in areas such as literacy, math, social
and emotional domains, and science, they need to be exposed to and engaged in intentionally designed
and implemented curricular experiences (Pianta et al., 2020b). It is beyond the scope of CLASS to measure
the quality of specific curricular experiences or content-specific instructional practices (see “What CLASS
Does Not Measure”). However, CLASS observations capture information about the number of children, the
amount of time children engage in content areas, and the format for instruction. This information enables
educators to understand how interactions and activity settings come together to enhance children’s
learning and development.
How was CLASS developed?

CLASS was originally developed based on an extensive review of literature and observational systems
used in large-scale studies such as the National Institute of Child Health and Human Development (NICHD)
Study of Early Care (NICHD ECCRN, 2002, 2005b; Pianta et al., 2002) and the National Center for Early
Development, Teaching, and Learning (NCEDTL) Multistate Pre-K Study (Early et al., 2005). The domains and
dimensions assessed by CLASS originated from a review of constructs assessed in classroom observation
instruments. Childcare and elementary school research, literature on effective teaching practices, focus
groups, and extensive piloting, use these observation instruments.
Educator-child interactions can be reliably described,

observed, measured, and improved on a large scale.
NCEDL studies, which included 2,400 children from 671 learning settings in 11 states, validated the Pre-K
and K–3 CLASS. The NCEDL studies found that
● Effective Instructional Support was associated with greater preschool gains in receptive
vocabulary, expressive vocabulary, rhyming skills, letter-naming skills, and mathematics skills
(Mashburn et al., 2008).
● Effective Emotional Support was associated with gains in social competence and fewer behavior
problems (Mashburn et al., 2008).
● CLASS predicts child outcomes equally well in learning settings with high enrollments of DLLs,
illustrating the validity of CLASS in these settings (Downer et al., 2011).
The domains, dimensions, and indicators of CLASS have proven valuable in measuring the effectiveness
of educator-child interactions. Research evidence and practice requirements are the basis of the additional
elements of quality measured in the second edition. As an example, there is somewhat mixed evidence
on the specific environmental contributions to children’s learning and development (McDoniel et al., 2022;
Soliday Hong et al., 2019). However, these details are a key element of quality to capture for educators
and educational systems. Many of the other updates in this edition reflect working with a group of experts
and educators from varied backgrounds, with a diversity of expertise. This work expanded descriptions of
quality to better reflect the learning settings in which young children spend their time. Chapter 5 provides
details on the approaches and outcomes of this work.
Other research has corroborated early findings on CLASS. These findings demonstrated that educator-child
interactions can be reliably described, observed, measured, and improved on a large scale in ways that
affect children’s social and academic outcomes. Chapter 2 summarizes this research.
Summary
The CLASS framework and the aligned measurement suite capture the quality of interactions in a learning
setting as a key component of CQI efforts in early education. CLASS 2nd Edition has strong theoretical
and empirical foundations in child development and uses the current enhancements to address equity and
additional components of the learning setting. The CLASS 2nd Edition will continue to provide essential
data on such settings to support educators’ effectiveness and children’s outcomes.
®
Chapter 2:
CLASS Research Summary
Since the publishing of the first papers on CLASS, a growing body of research has documented the ways
in which educator-child interactions support learning and development across domains, grade levels,
populations, and settings. Research also documents the ways in which intentionally designed professional
development and policy systems can lead to systematic improvements in educator-child interactions. This
chapter provides a brief overview of this literature and addresses some of the limitations of that research.
CLASS and Child Outcomes

More than 140 studies demonstrate that children make greater gains in key areas of school readiness—
including language, literacy, math, self-regulation, executive functioning, and social and emotional
development—when they experience more effective educator-child interactions. These studies often used
rigorous research methods, such as random assignment or assessment over time, that help to solidify the
nature of the findings. The following section outlines key findings across these areas. Briefs on CLASS
research can be found at teachstone.com.
Language and Literacy. Effective educator-child interactions, as captured by all three CLASS domains, predict
children’s language and literacy skills both within one school year and over the course of several school years.
Children who experienced warm and supportive relationships, engaging lessons, effective feedback, and
language modeling were more likely to perform better on measures of vocabulary, phonological awareness,
letter-word recognition, and reading comprehension (Bulotsky-Shearer et al., 2020; Leyva et al., 2015;
Mashburn et al., 2008). Children in Head Start specifically showed improved letter-naming and vocabulary
skills when in settings with higher levels of Instructional Support (Han et al., 2017). Children who were DLLs
showed increased skills in phonological awareness, letter-naming, and vocabulary in settings with effective
Emotional Support and Instructional Support (Hindman & Wasik, 2015; Neugebauer et al., 2020).
First-grade children who experienced higher overall interaction quality in kindergarten were more likely to
score well on tests of sight-word and decoding skills than first graders who had less effective interactions
(Lee & Bierman, 2016). Furthermore, children demonstrated stronger reading comprehension and literacy
skills at the end of third grade when they experienced more years of effective teaching, as defined by
CLASS (Vernon-Feagans et al., 2019).
Math. Improved educator-child interactions, as measured by CLASS, are associated with gains in
mathematics (for example, applied problems and quantitative concepts) in preschool (Finders et al., 2021;
Hu et al., 2016; Pianta et al., 2020b). Higher levels of Emotional Support in kindergarten are associated
with greater gains in applied problems (Johnson et al., 2016). Higher total CLASS scores in pre-K and
kindergarten are associated with improved math skills, with significantly greater effects for children who
experienced the highest-quality kindergarten (Carr et al., 2019).
Self-Regulation and Executive Functioning. All three domains of CLASS relate to children’s self-regulation
and executive functioning. Children who had difficulty with self-regulation were more likely to engage in
positive social interactions when learning settings had higher scores in Emotional Support and Classroom
Organization (Broekhuizen et al., 2017). More effective Emotional Support interactions are associated
with improvements in self-regulation and inhibitory control (Hatfield et al., 2021) as well as a decrease in
children’s level of the stress hormone cortisol, which can inhibit self-regulation (Hatfield et al., 2013).
Children who experienced more effective interactions also showed gains in social and cognitive skills, such
as inhibitory control and working memory, which are key to executive functioning (Hamre et al., 2014), as
well as higher general executive functioning skills (Leyva et al., 2015). Instructional Support interactions may
be particularly important in improving executive functioning skills (Alamos & Williford, 2020).
Social and Emotional Learning. Effective interactions play an important role in supporting children’s social
and emotional learning. More effective interactions are positively associated with increased inhibitory
control and social competence in preschool children (Choi et al., 2016; Weiland et al., 2013). Preschoolers
who experienced warm and responsive classroom interactions in the fall of the school year displayed more
prosocial behavior in the spring (Pakarinen et al., 2020). Similarly, when Emotional Support was higher,
elementary school children performed better on tests of social skills and showed decreased problem
behaviors (Broekhuizen et al., 2016; Burchinal et al., 2016).
These studies demonstrate a consistent link between the effectiveness of educator-child interactions and
children’s outcomes across various developmental domains. Additional evidence of links between CLASS
and child outcomes can be found at teachstone.com.
CLASS across a Diversity of Populations

A growing body of research conducted in programs within the United States, serving children from diverse
racial and ethnic backgrounds, describes the relationship between classroom quality as measured by
CLASS and children’s outcomes (Carr et al., 2019; Cash et al., 2019; Hamre et al., 2014; Vitiello et al., 2018).
For example, Hamre and colleagues collected data from 10 sites across the United States that served
racially and ethnically diverse children; almost half of the children in the sample were Black, one-third were
Latine, with smaller percentages of White, Asian, and other races or ethnicities. Data showed that children
who experienced higher-quality learning settings as measured by CLASS showed gains in social and
cognitive skills, including inhibitory control, working memory, and language and literacy skills (Hamre et al.,
2014). Furthermore, several studies show that children who are DLLs benefit socially and academically from
high-quality educator-child interactions (Limlingan et al., 2019; Hindman & Wasik, 2013; White et al., 2020).
Educators in all 50 states and in 50 other countries use CLASS in some form (such as for accountability,
professional development, or research). These countries vary in geographic location and economic
development and include Norway, the Netherlands, Australia, Chile, Ecuador, and Sri Lanka. Data strongly
suggests that the overarching three-domain structure of CLASS is applicable in cross-cultural settings (Hu
et al., 2016; Leyva et al., 2015; Slot et al., 2017).
Research from outside of the United States also suggests that interactions, as measured by CLASS, relate
to children’s outcomes (Cadima et al., 2010; Hu et al., 2021; Leyva et al., 2015). For example, a longitudinal
study conducted in Ecuador found that the way in which educators interact with children was most strongly
associated with their improved learning outcomes (Araujo et al., 2016). Similarly, a study conducted with
2,000 children enrolled in an Australian preschool suggested that higher levels of quality, as measured by
CLASS, led to better pre-academic skills. Gains were most pronounced for children with lower cognitive
skills (Niklas & Tayler, 2018). A longitudinal study in China demonstrated that Instructional Support
interactions predicted preschoolers’ growth in math and executive functioning skills in the second semester
of kindergarten. More effective instructional interactions were related to increased academic skills for
children (Hu et al., 2017).
Thus, research on the use of CLASS in a diversity of settings demonstrates CLASS can accurately measure
interactions in ways that predict children’s learning and development.
Improving the Quality of Interactions

Educator-child interactions are sensitive to forms of intervention and professional development designed
to improve them. Research has shown that targeted professional development helps educators improve
the quality of their interactions with children, leading to better child outcomes (Hamre et al., 2012b;
Mashburn et al., 2008).
CLASS-Based Interventions. Using CLASS in ongoing coaching helps educators improve their interactions
with children. For example, MyTeachingPartner (MTP) coaching—an intensive one-on-one CLASS-based
video coaching program—has been shown to improve the quality of educators’ interactions with children
(Allen et al., 2015; Downer et al., 2009; Hamre et al., 2010; Pianta et al., 2014). The initial MTP study showed
that the program led to improvements in educator-child interactions, with the most pronounced findings for
children in settings with high percentages of children in poverty. Furthermore, children whose educators
participated in MTP coaching displayed fewer problem behaviors and made greater gains in literacy (Hamre
et al., 2010). A statewide evaluation of three professional development models (including MTP) that focused
on the domain of Instructional Support found that, regardless of the model, educators improved their
interactions in this domain. Two groups also enhanced their interactions related to Emotional Support and
Classroom Organization (Early et al., 2017).
Targeted professional development helps educators

improve the quality of their interactions with children.
Other Interventions. Research also shows that professional development that does not specifically focus on
CLASS can support improvements in educator-child interactions. For example, a professional development
intervention focused on social and emotional learning moderated the relationship between an educator’s
self-reported burnout and their interactions with the children in their learning setting. Educators who
reported high levels of burnout decreased the efficacy of their Instructional Support across the school year.
In contrast, educators who participated in professional development did not demonstrate this decrease in
Instructional Support (Sandilos et al., 2020).
In another study, educators participated in an intervention to promote effective instructional practices for
young children considered to be at risk for developing an emotional or behavioral disorder. These educators
reported a greater sense of self-efficacy, and their settings scored higher in all three domains of CLASS
(Conroy et al., 2019). As one other example, researchers found that preservice teachers who participated in
a mindfulness intervention demonstrated higher-quality classroom interactions (Hirshberg et al., 2020).
Research Limitations
Although hundreds of studies report statistically significant relations between measures of quality, such as
CLASS, and child outcomes (Mashburn et al., 2008; NICHD ECCRN, 2002; Sabol et al., 2013), limitations of
educational and development research affect the interpretation of the results. Most notably, when researching
links between classroom quality and child outcomes, effect sizes (or the size of the relationship) tend to be
modest, and in some instances non-significant (Brunsek et al., 2017; Burchinal, 2018; Perlman et al., 2016).
Evidence from causal research designs that include random assignment of children to educators confirm that
educator-child interaction had small but significant positive causal effects on learning (Carneiro et al., 2019).
As the field considers ongoing refinement to existing tools and the development of new tools, it is important to
consider effect sizes in light of four realities: assessed child outcomes, multiple elements of quality, sampling
constraints, and reliability of data.
Outcomes. Nearly all studies of quality, including CLASS, use standardized assessments of outcomes
that are not directly related to the measures of quality. Such assessments (Wendling et al, 2015) capture
generalized performance or achievement and are not aligned to specific instructional experiences. Perhaps
it is not surprising that assessments of the overall effectiveness of interactions demonstrate modest effects
on general measures of outcomes. Interestingly, in the upper grades, relationships between CLASS and
outcomes tend to be greater. This may be a result of educators matching instruction to more specific
standards assessments required in these grades (Allen et al., 2013).
Multiple Elements of Quality. Evidence is increasing for a “package” approach to understanding quality. It
may be that measuring several aspects of quality is most helpful in understanding the relationship between
quality and child outcomes. This package likely includes educator-child interactions, dosage of content, use
of a targeted curriculum, and alignment of instruction to children’s skills (Nores et al., 2022). Such a package
predicted children’s learning and yielded a larger effect size than each individual component alone (Pianta,
Whittaker et al., 2020b). Quality, even when defined by direct influences on development, is multifaceted; it
is unlikely that any single quality assessment will measure all of them well. As related to CLASS 2nd Edition,
Teachstone recognizes that CLASS will only affect specific academic outcomes, such as in literacy and
math, if children experience that content within a strong curriculum. The inclusion of stronger activity setting
data may be useful to increase correlations to outcomes, as prior studies suggest.
Sampling of Quality. Some studies conducted CLASS observations in the fall and the spring whereas other
studies only collected data once a year. Regardless of whether CLASS observations take place once or
twice a year, observers are obtaining a very small “slice” of quality (typically, between two and four hours
over a full year). This makes it difficult to accurately predict gains in children’s learning and development.
More frequent data collection allows for greater accuracy in measuring outcomes. Innovations in machine
learning may offer advances that allow for more frequent sampling of the quality of interactions without
adding significant cost or time burdens (Ramakrishnan et al., 2021).
Reliability of Observers. Correlations between CLASS and outcomes are also affected by the reliability of
the data collected, with evidence of considerable observer or rater variance in some studies (Mashburn
et al., 2014). Although Teachstone requires a minimum annual recertification, ensuring that observers are
using CLASS effectively requires more rigorous efforts. Variability in observer supports may also reduce
associations between CLASS and outcomes by lowering the reliability of CLASS scores. Chapter 4 outlines
recommendations for supporting reliability. These recommendations include ways to support training and
calibration among observers and ways to randomly assign observers to minimize the impact of systematic
observer effects.
Summary
Taken together, the research on CLASS demonstrates several key learnings:
● Interactions can be observed reliably.
● Interactions are important predictors of child outcomes.
● Interactions can be systematically improved.
● Interactions can be measured at scale.
● The CLASS tool is an effective way to understand and support interactions.
Understanding how all the elements of children’s classroom experiences work together to support learning
and development, how to measure these on a large scale, and how to address inequities in classroom
interactions requires further research. CLASS has demonstrated significant progress in understanding the
kinds of interactions that predict child outcomes. However, innovations in the years to come can help further
refine the CLASS measure and facilitate development of new approaches to the measurement of quality
and children’s classroom experiences.
®
Chapter 3:
CLASS in Policy and Practice
In addition to use in research, CLASS is extensively used as a measurement tool within the United States.
The federal government, states, and localities require objective ways to measure the quality of early
childhood education and to ensure that investments in quality improvement are effective. The following
section briefly describes the use of CLASS by the Office of Head Start (OHS), state Quality Rating and
Improvement Systems (QRIS), and as a part of quality improvement systems in pre-K–12 settings. Although
CLASS was initially adopted in many of these settings as an accountability measure, it has evolved to
become a unique CQI system that allows for coordinated focus and investments in the assessment and
improvement of interaction quality at large scale.
CLASS and the Office of Head Start

Head Start has helped transform the definition of high-quality early childhood education by focusing
on those educator practices that make the most difference: educator’s interactions with children. The
2007 reauthorization of the Head Start Act authorized the use of a Designated Renewal System (DRS) to
improve the assessment of quality in Head Start programs. The DRS requires the federal government to
evaluate grantees in a new way, shifting the structure from grants funded in perpetuity to a five-year grant
program, through which grantees receive an additional five-year contract by demonstrating competence as
measured by the DRS. Under the DRS, programs that do not meet the quality requirements of the system
must recompete for a five-year award. CLASS was one of seven triggers that result in a program having
to recompete for funding. Initially, CLASS scores that fell below the thresholds established by OHS had to
recompete. In the fall of 2020, OHS removed the criterion that the lowest 10% of programs had to recompete
and established new competitive thresholds. The threshold for Emotional Support increased from 4 to 5 and
Classroom Organization increased from 3 to 5. The Instructional Support threshold increased from 2 to 2.3
until July 31, 2025. After that, an additional increase to 2.5 will go into effect.
Since the inception of the DRS, Head Start programs across the country have engaged in professional
development focused on classroom interactions, and grant recipients have made steady improvements.
As a result, there have been small but systematic improvements in the quality of classroom interactions for
Head Start children (Aikens et al., 2016). Such threshold changes imply that many programs were already
hitting the thresholds and indicate that improvement efforts and the use of CLASS are working. Programs
continue to increase their CLASS scores and provide more children with access to effective interactions
(OHS, 2018).
CLASS and QRIS

Since the late 1990s, Quality Rating and Improvement Systems (QRIS) have been widely adopted
to raise the quality of early childhood programs across the United States. The goal of QRIS is to
incentivize programs to improve quality through technical support, financial incentives, and publicly
available quality ratings. In 2022, most states and the District of Columbia either had an active QRIS,
were in the process of revising an existing QRIS, or were planning or piloting a QRIS. As of 2022, 23 of
these states included CLASS.
As policymakers seek to reform QRIS to address concerns about effectiveness and equity (Workman,
2021), several findings from the use of CLASS at the state level are notable. Sabol and colleagues
(2013) examined a variety of QRISs used by states for monitoring and evaluating program quality. They
found CLASS to be the strongest predictor of children’s outcomes when compared to other typical
QRIS metrics, such as educator education levels or educator-child ratio. Recent results from Louisiana’s
statewide system also show CLASS as a key component of predicting child outcomes (Markowitz et al.,
2020). Additionally, results from pilot efforts in Florida and Georgia indicate the use of CLASS as key to
improvement efforts (Early et al., 2014; Rodgers et al., 2018, 2020). These studies highlight that large-
scale implementations of CLASS help educators understand and improve interactions across settings and
age groups.
Systematic Improvements in Quality across Sectors

Despite all the investments in QRIS, there is inconsistent evidence that every component of the systems
links to child outcomes or leads to system-wide improvements in quality (Markowitz et al., 2020; Soliday
Hong et al., 2019). However, these studies show that key components of QRISs, such as the quality of
classroom interactions, demonstrate positive outcomes for children and programs.
The focus on interactions as measured by CLASS within a QRIS has led to effective and systematic
improvement in quality at the state level. Of note is work in Louisiana, one of the few states to develop a
mandated statewide QRIS across sectors. Louisiana’s unified early childhood system uses CLASS as the
sole measure to determine quality levels and targeted support. Using a systems-level focus across center-
based childcare, Head Start, and pre-K learning settings ensured that resources and capacity-building
focused on the same goal. That focus has improved classroom quality across all sectors since 2015. CLASS
scores, as well as buy-in from program leaders and educators, have increased steadily over the years. The
increase in scores has been especially evident in center-based and Head Start programs, which initially
lagged behind the state pre-K classrooms (Bassok & Markowitz, 2020).
CLASS implementations at local levels have similar findings. For instance, Ventura County (within California’s
QRIS) partnered with Family Child Care Home and state preschool educators to drive improvements.
Over the course of two years, programs engaged in coaching on CLASS. At the end of the pilot, program
scores increased for these educators. Although this study focused on only two sectors of childcare within
California, the study demonstrates that CLASS supports improvement within various components of a
state’s QRIS (Teachstone, n.d.).
CLASS indicates where additional resources can be most

impactful in ensuring every child has access to effective interactions.
Ensuring More Equitable Distribution of Resources

Within a QRIS, states often create tiers or levels to indicate quality, as determined by various criteria such
as education levels, educator-child ratios, curriculum used, and observed quality. These levels are generally
tied to amounts of funding to incentivize program improvement; programs with higher levels of quality
receive higher subsidy rates. This method of funding leads to inequity in the allocation of resources, as it
can be difficult to reach and maintain high levels of quality without adequate funding. Rather than wait until
programs achieve high-quality ratings to provide resources for improvement, CLASS highlights strengths
and opportunities for growth at all levels of quality and indicates where additional resources can be most
impactful in ensuring every child has access to effective teacher-child interactions.
An example of such a system is Florida’s Early Learning Performance Funding Project, a pilot designed
to inform Florida’s School Readiness Program (Florida Department of Education, n.d.). Implemented in
2014, and involving almost 3,200 educators, this project required CLASS observations to make decisions
about training, compensation, and other supports. The project assigned quality tiers based on CLASS
ratings. After determining a quality tier, providers had flexibility to select a CQI strategy to best meet their
program’s needs. Providers in Tiers 1–3 needed to complete specific quality improvement strategies.
However, providers in Tiers 4 and 5 could complete multiple options but had no requirement to engage in
professional development. Programs then participated in research-backed professional development and
CQI interventions, such as CLASS Group Coaching—formerly known as Making the Most of Classroom
Interactions, or MMCI. A multi-year project evaluation of this pilot showed that engaging in the professional
development and using CLASS for CQI efforts provided the foundational knowledge and skills that first-tier
providers needed to improve educator-child interactions (Rodgers et al., 2018). In fact, the first-tier programs
showed the most growth in interaction quality over the course of the implementation when compared to
the other tiers (Rodgers et al., 2020). This evaluation also showed that the combination of improvement
strategies and support strategies at each quality level produced gains in CLASS. The evaluation also
showed that program improvements related to other aspects of quality, such as increased professionalism,
improved communication with children and families and improved collaborative leadership among directors
and educators (Rodgers et al., 2018).
Providing Data to Support CQI

Improvements for any organization require frequent access to data on current performance. Yet, a QRIS often
● Relies on infrequent or incomplete collection of data to calculate ratings
● Uses complex rating calculations that may mask performance on the individual quality components
● Uses program scores that do not provide individual educators with the information they need to make
improvements
When systems function in this way, rather than through a transparent CQI process, it is harder for programs
to use data to focus on improvements.
In Louisiana’s statewide CLASS implementation, consistent feedback and increased transparency for
programs were prioritized to help ensure those closest to children and families had what they needed to
improve children’s experiences (Bassok & Markowitz, 2020). Louisiana required that every learning setting
in each site that was receiving public funds must be observed twice annually by community observers,
and that these results be entered into an accessible portal. Third-party independent observers conducted
additional observations and entered them into the same portal. Using such a portal, programs have real-
time access to information about their quality, and educators receive quick feedback. Communities can also
focus their improvement and technical assistance efforts to support issues, concerns, and trends in a timely
manner. This comprehensive data collection system resulted in a clear understanding at all levels about
what quality looked like at the classroom level. The data also showed where additional resources needed
to be targeted and what goals for improvement should be (Bassok & Markowitz, 2020). Regardless of the
specific system, educators need regular and clear access to the data from their observations to engage in
CQI efforts.
Investing in High-Quality Improvement Programs

QRISs that focus only on the rating element, without attending to support for high-quality professional
learning opportunities for educators, are unlikely to lead to systemic improvement. States and localities
have thoughtfully implemented elements of the CLASS CQI System to help support improvement on a large
scale. Examples from Louisiana and Florida (see previous sections) made use of many Teachstone products
and services, most notably CLASS Group Coaching (MMCI) and CLASS 1-on-1 Video Coaching (MTP).
Georgia conducted a more rigorous evaluation of similar work.
Between 2011 and 2014, the Georgia Department of Early Care and Learning provided two forms of
CLASS-based professional development for lead educators in their state-funded pre-K program, Bright
from the Start (Early et al., 2014). Both programs had the goal of increasing educators’ understanding of
effective educator-child interactions, and of increasing the intentionality of these types of interactions.
Educators were randomly assigned to one of two forms of professional development–MMCI or MTP–or
a control group. MMCI consists of 10 sessions of face-to-face professional development for a group of
educators. MTP is an intensive one-on-one video coaching program that pairs educators with virtual
coaches over the course of a school year. CLASS observations conducted in the fall and spring of each
year showed findings from learning settings with educators who participated in MMCI. These educators
had demonstrable gains in both Emotional Support and Instructional Support scores when compared to
those in the control group. Similarly, educators who received MTP coaching had learning settings that
scored higher in Emotional Support compared to the control-group educators (Early et al., 2014). These
findings highlight that both types of coaching products and services support educators in improving
interactions within a larger system. They provide systems with options for how to include CLASS-based
coaching within their professional development.
CLASS in Pre-K–12 Systems

Schools across the country are continuously looking for ways to improve children’s learning and
development. Measuring and improving instruction is a major focus of this work. CLASS has been used less
often by pre-K–12 systems as an accountability tool. However, some research exists on CLASS as a tool
for educator evaluation and school improvement in pre-K–12 systems. CLASS is more often used as part of
the districts’ CQI efforts. Studies show that the use of CLASS in pre-K–12 systems supports more effective
interactions and shows links to child outcomes in each case.
CLASS and Educator Evaluation

The Measures of Effective Teaching Project, funded by the Bill & Melinda Gates Foundation, compared five
frequently used classroom observation tools. This study helped educators and policymakers identify and
support teaching by improving information about what constituted good teaching. In addition to conducting
educator observations, the project looked at gains in student achievement and student perception surveys.
CLASS was one of two measures selected to assess overall instructional quality. The study used the Upper
Elementary and Secondary CLASS tools, along with the Framework for Teaching (FFT), to assess educator
effectiveness in 3,000 classrooms (Danielson, 2014). Findings suggest that although no single measurement
tool provides sufficient information to evaluate teaching quality, of the two observational measures,
CLASS consistently had higher correlations with student outcomes (Kane & Staiger, 2012). This research
also showed the ways in which pairing classroom observations with additional measures, such as student
surveys, can provide a more robust assessment of teaching quality—at least at the older age levels.
CLASS and School Improvement

The Massachusetts Department of Elementary and Secondary Education partnered with the American
Institutes for Research (AIR) to evaluate methods of assisting low-performing schools—referred to as
“turnaround schools”—in improving school quality. They worked in elementary and secondary schools
that served a greater percentage of Black, Hispanic, DLL, and economically disadvantaged children than
the state as a whole. One part of their efforts included regular CLASS observations, as often as once a
week, with targeted feedback on educators’ instructional practices. At the end of the project, standardized
testing showed remarkable results among the children in the lowest-performing schools (determined
by achievement, child growth, and improvement trends captured by the Massachusetts Comprehensive
Assessment System [MCAS]). These children made gains in both math and English language arts that were
equivalent to an additional year of instruction (Stein et al., 2016).
CLASS and District CQI
Dallas Independent School District (ISD) has been using CLASS to measure quality in pre-K classrooms
since 2015 and over time has added kindergarten and first- and second-grade classrooms. The district
created a culture of CLASS by investing in its coaching program. Evaluations showed that children who were
in a high-quality pre-K classroom as measured by CLASS were better prepared for kindergarten than their
peers who either did not attend pre-K or experienced low levels of quality (Wright et al., 2018). Children who
were in a high-quality classroom in pre-K and kindergarten were more likely to be on track in first grade
than their peers in less effective classrooms. Furthermore, data revealed that experiencing multiple years
of high-quality instruction was the best way to help ensure that children were set up for success in the
following school year. Additionally, Dallas ISD has seen a steady overall increase in CLASS scores over the
years, due in large part to the district’s focus on assessment and professional development efforts aligned
to effective interactions. Only 40% of their preschool classrooms scored at or above 3.25 in Instructional
Support in 2015. This percentage increased to 60% in the fall of 2017 (Wright et al., 2018).
Summary
Overall, these studies of national, state, and local systems demonstrate that large-scale implementations
of CLASS can be highly effective in improving interactions and other outcomes for programs. Specifically,
Head Start’s consistent use of CLASS in measurement and improvement efforts show that programs can
continue to show gains over several years of implementation (Aikens et al., 2016). The work in Louisiana
demonstrates that the more community and program leaders know and understand the QRIS, the more
they can buy into the CQI process and in turn support improved child outcomes through their work
(Bassok & Markowitz, 2020). As the work in Dallas ISD highlights, building local capacity and expertise
around the vision of quality for the system is also an investment in the sustainability of the system and
the experiences that children have across grades (Wright et al., 2018). Additionally, the work efforts in
Florida and Georgia demonstrate that providing support and resources for programs, especially for those
that have historically been under-resourced, also leads to significant gains (Early et al., 2014; Rodgers et
al., 2018, 2020). Programs, educators, and children benefit when systems can use data and resources to
provide focused support.
Chapter 4: ®
Establishing a CLASS Observation

and Reporting System
This chapter addresses general considerations for establishing key components of a CQI system, including
information on developing effective protocols to obtain CLASS data, recruiting and training observers,
ensuring data quality assurance, and providing data and feedback to educators. The Observation Field
Guide contains specific information for how observers should conduct themselves before, during, and after
observations, and how to score the observations. As data collection and reporting systems involve multiple
components, more detailed information to establish an effective CLASS observation protocol may be found
at teachstone.com.
CLASS Observation Components

A CLASS observation generally consists of between one and six cycles, depending on the purpose of the
observation, and the data-sharing and decision-making levels (educator, classroom, program, school).
Each cycle typically consists of a 15–20-minute observation period and about 10 minutes for coding.
At the end of each observation period, the observer will
1. Stop observing the learning environment.
2. Spend about 10 minutes assigning indicator ranges, dimension scores, and Activity Setting item
ratings using the guidance and coding process outlined in the Observation Field Guide.
Then the observer will begin a new observation cycle and repeat the process until they complete the
required number of cycles. For information on how to conduct the observation and assign scores, including
the Activity Setting items, see the Observation Field Guide.
Supporting Readiness in the Organization

Before starting CLASS observations, everyone within the organization must understand why CLASS data
collection is happening and how it will affect their work. This is a critical part of the “Focus” stage of CLASS
implementation (see Figure 1.2). Ideally, leadership will develop a CLASS implementation plan that helps
ensure everyone in the organization understands how the use of CLASS aligns to broader program goals
and how it will help them in their daily work.
Individual and systemic buy-in and readiness affect the success of an implementation plan and what
messages best support continuous improvement efforts. Educators who are ready to engage in CQI see a
need for improvement and feel capable of making changes with support. Leaders can help educators get
ready for CQI efforts by ensuring they understand
● How more effective interactions lead to real improvements in children’s experiences and outcomes
● What those effective interactions look like
● That they can improve their interactions with children
Reviewing these messages throughout the CQI process will keep motivation and engagement high.
Teachstone strongly recommends that all staff members review information about CLASS prior to the start
of observations. This should include an overview of the CLASS framework and role-specific information
about its use. Educators must also know what to expect during the observation and how and when they will
receive feedback.
Developing an Observation Protocol
Prior to collecting CLASS data, set up a clear observation protocol that details the steps observers need
to take during an observation. Doing so helps ensure that observation data is reliable (accurate) and valid
(measures what it is designed to measure) across all learning settings and observers. Having reliable and
valid CLASS data helps ensure that organizations have the information they need to best support educators
and children in their programs. Teachstone provides high-level guidance on CLASS Observation Protocols
while also recognizing that the protocols will vary depending on the needs of the organization. The standard
protocol for CLASS observations, detailed in the following section, finds a balance among feasibility, cost,
and accuracy of data. However, this protocol can vary depending on the purpose of observations (see
“Adapting CLASS Observation Protocols to Align with Purpose”).
Standard CLASS Observation Protocol

The following information highlights key aspects of a standard observation protocol in terms of when and
what to observe.
When to Observe
Observe in fall and spring. Teachstone recommends that formal observations be conducted every
year in the fall and the spring. This allows programs to determine the effectiveness of their professional
development programs or other CQI efforts. This schedule also allows program coordinators to look at
growth year over year.
Avoid days right before or after holiday breaks and the first and last weeks of school. To obtain a more
realistic picture of a learning setting, educators and children must have time to settle into a new school year
prior to doing CLASS observations (about three to four weeks). CLASS scores tend to be slightly lower in
the first few or last few weeks of the school year or the days around holiday breaks. Avoid these times, if
possible, to ensure scores reflect a typical day.
Begin observing at the beginning of the day or class. When observing in learning settings where children
stay with the same educators for most of their day, the observation should start at the beginning of the day
and continue throughout the morning until the observer completes the predetermined number of cycles.
Schedule observations. When scheduling an observation, the observer should obtain the daily schedule to
determine the appropriate times to begin and end the observation. This allows them to maximize the chance
of completing the desired number of cycles. Scheduling is particularly important in preschool classes, where
children may spend significant amounts of time playing outside. It is also important in elementary classes
in which children may attend “specials” or a course taught by a different educator, such as music or PE, as
these activities may not be codable.
What to Observe
Conduct a minimum of two hours (four cycles) of observation per day. If the goal is to share formal CLASS
data with educators at the classroom level, Teachstone recommends at least two hours of observation
(typically four cycles). Observation periods are typically 15–20 minutes in length, followed by about 10
minutes of scoring. These cycles then repeat over the two-hour period. Scores should not be given for
observation periods of less than 10 minutes.
Pause scoring during times that are deliberately unfacilitated. CLASS focuses on observing the quality
of educator-child interactions, inside or outside, during more formal and informal moments. However,
observers should pause scoring during times that are designed to have limited opportunity for interactions,
such as recess or testing. Observers should score center time and all other activities in which children
engage in work and play. Additionally, long independent times (for example, Montessori work periods, older
children completing an independent writing assignment) are codable. At these times observers should look
for the educators to maintain awareness of the entire learning setting and may score interactions that occur,
even with just one or two children.
If recess or testing occurs during an observation, observers should stop scoring unless the observation
period has been less than 10 minutes. If the cycle needs to end before 10 minutes of the observation period
have occurred, the observer may not assign scores and must start a new cycle once activities resume. It is
fine, and in fact desirable, to observe during transitions and other routines.
Adapting CLASS Observation Protocols to Align with Purpose

The standard CLASS observation protocol described in the previous section works for many CLASS
implementations but can be flexibly adapted to meet the needs of the organization. This section
provides guidance on some of the most common adaptations. It is important to remember that even the
recommended two hours of observation twice a year in the standard protocol is a relatively thin slice of a
learning setting. Longer or more frequent observations provide more accurate data on children’s typical
experiences. If resources are available for more regular observations, observers can score each learning
setting more frequently rather than for a longer period each day. There is greater day-to-day variability in
CLASS scores than variation across a given day. More frequent observations also provide educators with
additional opportunities for feedback within a CQI system. Table 4.1 outlines the most common purposes for
observation and how specific aspects of the protocol and system differ based on the purpose.
Areas of Adaptation
The following are additional considerations and recommendations for what and how to observe to meet the
needs of the organization.
Observing a sample of classrooms. Observing every room so that all educators in the program have
equitable access to data for CQI purposes is ideal. However, if the primary purpose is to gather summary
data at the program level, a sampling of rooms may be sufficient. The number of rooms needed for the
sample is highly dependent on the specific aims of a CLASS data collection. However, data from a program
-level sampling should not be shared with educators whose rooms were not observed. The information will
not be representative of their individual classrooms. These data should only share program-level trends or
be shown to the educators being observed.
Observing more or less frequently each year. In some situations, it may be difficult to observe all learning
settings twice a year. If this is the case, conducting observations once a year is acceptable. However, to
compare across years most accurately, collect data at a similar time of year. Similarly, in some research
projects or intervention evaluations, project administrators may opt to conduct CLASS observations once at
the beginning of the project and once at the end.
Observing for fewer or more than four cycles. If data collected is for evaluation or accountability purposes
only at the program level (averaged across classrooms), and not being shared with individual educators, that
may warrant fewer cycles. In general, there is more variability between classrooms in a program than there is
across a day in a single classroom. Therefore, observing more rooms for less time (for example, two cycles)
is likely to lead to more accurate program-level data. In other words, reducing the amount of time spent in
each room allows observers to visit a larger number to capture a more realistic or “truer” picture of how
educator-child interactions look at the program level. However, if decision-making is based on CLASS data at a
classroom level, conducting at least four cycles will provide a more accurate representation of interactions for
a specific learning setting.
For educator-level professional development, the observer may wish to conduct a full four-cycle
observation at the beginning of the year, followed by one- or two-cycle observations to obtain concrete
examples of interactions to share with an educator. Additionally, some organizations require a six-cycle
Table 4.1. Variations for Observing Based on Purpose
When to Who to How long

Purpose Considerations
observe observe to observe
Assess general Beginning of All educators 4–6 cycles Observing less than 50% of the
classroom the day when children may affect the types of
All children 15–20 minutes
quality at least 50% of interactions typically seen with
each
children have more children present.
arrived
Assess quality During time All educators 4–6 cycles Observe fewer cycles across
of specific when the teaching that multiple days to gain a clear
15–20 minutes
content area educator subject picture.
each or full
instruction presents
All children lesson if 30
specific
minutes or less
content
Learn the When the All educators 4–6 cycles May need to observe children
experience cohort is with more than one educator and
Children in the 15–20 minutes
of a cohort of together observe across multiple days.
cohort each
children
May follow the cohort as they
move into different subject areas
or with different educators (such as
for library or music).
Focus on When that Educator of 4–6 cycles May need to observe more than
a specific individual is focus and once to obtain the needed number
15–20 minutes
educator or teaching children they of cycles.
each
preservice work with
teacher
Assess at During a All educators 2–4 cycles If observing all learning settings
the program similar time All children is not possible, observe more
15–20 minutes
level (not the across learning learning settings for less time.
each
classroom settings, if
level) possible
observation to obtain more information about children’s experiences during the school day, such as
“specials” or afternoon activities.
Observing for longer or shorter cycles. Twenty-minute cycles help balance the goal of observing and
scoring sufficient instruction while not requiring observers to remember what they have seen over a long
period of time. Observation cycles should not be shorter than 10 minutes because the information is not
sufficient to score every indicator and dimension within that time. However, it is possible to plan for shorter
cycles (for example, 15 minutes) if that helps complete four cycles of observation. Longer observations may
be useful in some situations. For example, if a class session or lesson lasts for 30 minutes, it may be more
useful to score the full 30 minutes to provide feedback on a more holistic part of the instructional time.
Observing only one educator. Some professional development and research purposes require obtaining
ratings that reflect one educator rather than the entire setting. Observers should still follow the general
procedures described in the standard observation protocol in most cases. The main educator is often
responsible for directing children’s experiences. Therefore, information about the interactions that children
have with other adults may be important indicators of how the primary educator has structured the setting.
Some exceptions to this include, for example, work with preservice teachers. The scoring of these educators
focuses specifically on their interaction with the children rather than the interactions between children and
the primary educator. Therefore, if the preservice teacher is only working with a small group of children
throughout the observation, the observer should code only those children’s experiences.
Observing outside. Observers should not code outdoor recess or other outdoor unstructured “free
time” because these periods provide limited opportunities for such interactions. However, they may code
structured outdoor activities such as an educator-led nature walk, a science experiment, or any other
lesson that takes place outside. Note that some schools spend many lessons or daily activities outside. The
observer would code everything that takes place outside in these instances, except for a true recess period.
Observing in “specials.” The decision regarding whether to observe during specials will depend on
the purpose of the observation. If the purpose is to obtain information about the main educator(s), the
observer would not conduct observations during classes taught by other educators (art, computer,
physical education). On the other hand, if the purpose is to learn about the overall experiences of the
children in the learning setting, the observer would collect data during specials in the regular room or
in another room (for example, physical education in the gymnasium). The same guidance applies to
observing at times when other adults, such as an administrator, volunteer, or regular visitor, interact with
children during an observation.
Observing during therapy or intervention. Specialized instruction such as speech and language therapy
or intensive reading intervention have traditionally been delivered outside of the general education setting.
When individual children go to a different room for this type of instruction, the CLASS observer generally
stays with the main room and focuses on the remaining children. However, if a cohort of children leave
the room together, and the purpose of the observation is to gather information about how that group
experiences their time in school, the observer may follow them.
In contrast, for services delivered within the general education setting, the observer should take the
specialists’ interactions into account. The observer should score the specialist based on the following
principle: weigh their interactions according to the number of children with whom they are working, the
amount of time they spend with the children, and their responsibility for the activities. In some instances
the specialist may work with one child at a time, which would reduce the overall impact of their interactions.
However, there may be times when the specialist works with a group of children or even leads a whole-
group activity. In those circumstances their interactions would weigh more heavily. If the focus of the
observation is to gain information about the main educator, the observer would not include the specialists’
interactions in the overall score.
Announcing observations. If the objective is to observe specific classes or activities, scheduling the
observation in advance will ensure that schedules align with the intent. However, if classrooms are randomly
selected from the total number in the program, organizations may decide to use an observation window. For
example, a data collection administrator might decide that an observation will take place during a specific
week or two-week period. Observers or the scheduling coordinator should ask their point of contact
(program director or supervisor) to let them know any specific dates and times within that window that
would not work (for example, assembly, field trip, class party). Either approach can work if communicated
clearly to program administrators and observers.
Observation Method
CLASS has been validated for use in coding video of classroom activities in addition to live observation
(Curby et al., 2016; Farley et al., 2017; Guo et al., 2010). This method has been widely used in numerous
research studies and in an increasing number of field applications. The standard protocol and its potential
adaptations are applicable, regardless of observation method. The following section outlines logistical
recommendations for each method.
Live Observation Procedure

Before and After Live CLASS Observations. Learning about the setting prior to an observation is a practice
for reducing bias (see Chapter 5 for more details). When entering the room, observers should introduce
themselves to the educators, confirm they are in the correct room, and explain the observation process
(for example, they will take a lot of notes and move around the room to see or hear better). If educators
are already engaged with the children, observers may wait a few minutes to check in with the educators.
Following the observation, observers should thank the educators for their time and complete any follow-up
procedures as outlined in their specific observation protocol.
During Observations. Observers should review the Observation 2nd Edition Field Guide for specific
guidance on how to observe and score. Additionally, observers should follow any project-specific protocol.
To minimize disruption, observers should
● Stand or sit in an inconspicuous place to observe, such as near the wall or in a center that is not being
used, moving as needed to see or hear better
● Limit interactions with children or educators
Where to Code. The decision whether to stay in the room or step outside to code will depend on the
circumstances, the project-specific protocol, and the physical setup of the room or setting. Factors to
consider include the observer’s ability to focus on coding in a busy environment, whether leaving and
reentering the room would be disruptive to the children, and the availability of a nearby location for coding.
Regardless of where the observer codes, they should ignore interactions that occur during the coding
period. After assigning their codes, the observer should begin a new cycle until they have completed the
required number.
When to Stop an Observation Cycle. If the class moves into an activity or time that is not codable according
to the project-specific protocol or guidelines (see “Adapting CLASS Observation Protocols to Align with
Purpose”), the observer should stop observing. Additionally, if a routine interruption such as a fire drill
occurs, the observer should stop observing. If the observation period is more than 10 minutes long, the
observer can code that period and count it as a cycle. If the observation period is less than 10 minutes
long, the observer should not code that time and should start a new cycle when the activity is considered
codable. Observing and coding transitions to and from such activities, such as walking to recess or cleaning
up materials prior to a test, may count in the time for the observation period.
If an incident occurs that would require mandated reporting, the observer should make note of the incident,
stop the observation, and follow their project and state procedures for mandated reporting.
Video Observation Procedures

The general procedures described for live coding also apply to coding from video. This section includes
some additional considerations for video coding.
Coding Using Video

Observers should refer to project-specific protocols for when to start and stop the video (for instance, starting
the next cycle at the immediate ending point of the previous cycle or selecting specific cycle start and stop
times to align with another observer). Here are a few considerations to remember when using video:
● Use headphones to improve audio quality.
● Ensure access to high-speed internet to view content without interruption.
● Do not replay or rewind the video unless a technology issue in viewing occurs.
● Only code activities and interactions observed on-screen; do not factor interactions happening
offscreen into the scoring.
● Limit the number of cycles coded in one sitting to prevent coder fatigue. Teachstone recommends
breaking after four cycles and limiting daily coding to between six and eight cycles.
Collecting Video
To ensure the highest-quality data and consistency across observation methods, a video must adequately
capture the visual and auditory information present in interactions. The following are recommendations for
obtaining high-quality video:
● During whole- and small-group time, use digital video cameras that adequately capture sound without
requiring educators to wear microphones. However, in some cases such as group work or center time,
educators may be difficult to hear over the hum of activities. If a microphone is available, it can be
helpful in capturing sound. Experiment with a few options before recording a lot of video footage.
● Make sure the educator (or person doing the taping) tells the children what is happening prior to the
first video recording. The children should be told the reason for the recording and allowed to share any
concerns they may have about the process. Although children sometimes get excited about, become
distracted by, or shy away from the camera initially, most quickly forget it is there.
● If observing a specific activity or content area, start the video prior to the beginning of a lesson,
and record during the transition from one activity to another. These transitions are codable, and
often these non-lesson times provide interesting moments for educators to watch and discuss in
professional development.
● If observing a whole morning, start the video as soon as the camera is set up and the predetermined
percentage of children have arrived. Stop the camera when the needed number of cycles is complete
(generally one to two hours), or if an activity is not codable.
● Use a tripod and place the camera in such a way that the educators and most of the children are clearly
visible. Often it’s most helpful to set up the camera to the side to capture the facial expressions of the
educators and children. Move the tripod as needed, but do not use a handheld recording device, as the
video tends to be shaky.
● In the case of group work or centers, focus the video on the educator(s) leading the activity or engaging
more directly with the children, but occasionally pan out to capture the experiences of other children for
several minutes at a time.
● When collecting video data, follow the video consent, storage, and security processes specific to the
program, organization, district, or research institution.
Conducting Informal CLASS Observations

Routine informal observations are a key part of using CLASS as a CQI tool, as they enable more frequent
and targeted feedback for educators. Unlike formal observations that require at least a two-hour time
commitment, informal observations can be thought of more like “check-ins” or “walk-throughs.” Their
purpose is to gauge how well educators are progressing toward their CLASS-based professional
development goals and to provide actionable and specific feedback about progress and areas for continued
focus. During a formal observation, the observer is often a coach or a site-based leader who does not focus
on all 10 dimensions but rather on those identified as areas for growth. Because informal observations allow
coaches and leaders to provide real-time actionable feedback to educators, they can happen much more
frequently than formal observations. The observer does not assign scores but instead provides descriptive
examples of effective practices while also identifying growth opportunities related to the areas of focus.
For example, after an informal observation, the observer may share their feedback with the educator and
discuss the observation. They then jointly create an action plan that serves as a roadmap for professional
development. Many methods for conducting informal CLASS observations are available, and more
information can be found in the CLASS Observer Village and on teachstone.com.
Conducting CLASS Environment Observations

The CLASS Environment measure supplements CLASS observations with information about how to focus
on, measure, and improve environmental supports for high-quality educator-child interactions. Together, the
environment and interactions can enhance children’s learning and development. The CLASS Environment
is closely connected to the research-based domains of the CLASS tool—Emotional Support, Classroom
Organization, and Instructional Support—to organize measurement and reporting on the spaces in which
children learn. Items include key arrangements (for example, for cooperation and other purposes), displays
(for example, those that support routines and transitions or exhibits of children’s work), and materials (for
example, representation of diversity and items for imaginary play) that can support interactions.
A CLASS Environment observation is meant to seamlessly integrate with a CLASS observation. Teachstone
developed the CLASS Environment supplement for use alongside the CLASS tool. For the most meaningful
data, Teachstone recommends that observers collect CLASS Environment data while they conduct a full
CLASS observation. While the CLASS Environment adds the most value as supplementary CLASS scores,
there may be times when programs decide to use just the CLASS 2nd Edition to observe with more narrowly
focused goals in mind.
Upon arriving, observers should look around the room and spend about five minutes registering as many
environmental items as they can see before beginning the CLASS cycles. Observers should score what
they see as they move around with as little interference as possible. For example, an observer may move a
child-accessible material that is partially concealing another, as in the case of two books stacked together,
or remove the lid of a low storage container to register the materials inside. But observers should avoid
seeking out inaccessible materials that would require them to open cabinets, examine the contents of bins,
count items, or interview the educator.
Upon completion of an environment observation, the observer should complete their predetermined CLASS
observation cycles. For some environmental items, observers may need to wait until they have begun an
observation cycle (to see, for example, if and how children use technology or adaptive materials) to initiate
scoring. After completing a CLASS observation, the observer will take another five minutes to finish scoring
items that they did not observe or were unable to complete at the beginning of the observation. For specific
information on how to use the CLASS Environment, refer to the CLASS Environment Pre-K–K Manual. Note
that the CLASS Environment measure does not currently go beyond kindergarten.
Maintaining a Pool of CLASS Observers

A CLASS observer’s most important qualification is the ability to observe and code learning settings
objectively through the CLASS lens. An observer with a deep understanding of the CLASS observation tool
is more likely to collect accurate data and to provide supportive feedback to educators within a CQI system.
The following section outlines how organizations can recruit, train, and support CLASS observers to ensure
quality data collection.
Recruiting CLASS Observers

When looking for observers, it is best to select those with some background in education, as they may
be more familiar with the needs of children within an educational setting. However, all observers have
their own perspectives and preferences on what constitutes “good” instructional practice. They must put
their biases aside and learn how to code objectively using the CLASS lens. Observers must be able to
question and acknowledge their biases, learn strategies for recognizing and reducing them, and apply
those strategies during observations. More information on recognizing and reducing bias during CLASS
observations can be found in the Observation Field Guide and additional training resources.
Organizations should work to recruit a group of observers who represent the diversity of the programs
where they will observe. Given that it is beyond any single person to know and understand all possible
variations in settings, having a diverse group of observers will help ensure that they have real-world
experience and expertise in the settings where they will observe. They will also be able to more
accurately capture nuances in behaviors and interactions that may be less obvious to someone with
limited experience in that setting. Observers must fully understand the primary language(s) of instruction
to observe effectively.
Observers must be able to acknowledge their biases, learn strategies

for reducing them, and apply those strategies during observations.
Training CLASS Observers

A Certified CLASS Observer must complete each CLASS observation. To achieve certification, individuals
must attend training to learn about the CLASS framework and practice coding with the tool using video.
Observers then must take an online test to demonstrate their CLASS knowledge and skill. The Teachstone
website has a reliability support page to help ensure observers have what they need as they prepare for
certification or recertification. A schedule for these trainings is available at teachstone.com/trainings. All
observers must complete annual recertification to demonstrate they have maintained sufficient CLASS
knowledge and skill to use CLASS to accurately observe and score learning settings.
Once certified, observers must attend a project-specific orientation that provides the protocol for the project
and addresses common data collection scenarios, such as what to do if they cannot obtain the desired
number of cycles or educators are absent.
Observers should learn as much as they can about the environments in which they are observing. The more
that observers understand about the individuals and settings in which they observe, the better able they are
to notice and assess interactions that are supportive of children’s learning and development in the specific
context. Observers should obtain additional experience and training to continue to refine their skills prior to
observing and scoring.
Observers should also learn how to enter and maintain CLASS data as well as any plan to audit the
data once collected. Data may be recorded on paper score sheets or through Teachstone’s platform,
myTeachstone. The myTeachstone platform includes several measures to reduce data entry errors.
Supporting CLASS Observers

Certification is an important prerequisite for using CLASS. However, observers will benefit from additional
opportunities to refine their CLASS knowledge and to ensure they are following the recommended
practices for their specific data collection. Experienced observers within local organizations, and specific
Teachstone representatives, can provide this support.
In-the-Field Questions. Organizations should ensure that a team member certified on the CLASS tool and
knowledgeable about the protocol is available to answer questions that may arise while observers are in
the field.
Group Meetings. Administrators of the data collection efforts should schedule regular group meetings to
allow observers to discuss any coding-related or logistical challenges, hear updates on project processes, and
debrief calibrations. The frequency of meetings will vary by purpose. If an organization is conducting high-
stakes observations, Teachstone recommends more frequent meetings (biweekly or monthly). Observations
focused on professional development may require fewer meetings; observers may meet monthly, quarterly, or
at the beginning and end of the data collection window.
Teachstone-Specific Supports. The CLASS Learning Community is an online group that allows observers to
network with other CLASS observers and educators, participate in specialized discussion groups, access up-
to-date content, and ask questions. Similarly, Teachstone’s Observer Village is an online space full of tools and
resources to help observers build data collection skills, avoid coder drift (a decrease in reliability), and prepare
for recertification. In addition certified observers have year-round online access to Teachstone’s Video Library,
which contains examples of high-quality teaching and helps observers hone their skills. Teachstone also offers
Observation Support trainings on topics such as reducing observer bias and observing in settings with dual
language learners and can provide services related to data quality checks as further resources for observers.
Data Quality Checks

Accurate CLASS data depends on certified and well-trained observers, and an organizational commitment
to maintaining a high degree of fidelity. Data quality checks ensure that observers are coding reliably in the
field and help to prevent coder drift. These data quality checks include methods for assigning observers,
fidelity checks, double-coding, and calibration. The sections that follow contain descriptions of each data
quality check. The intensity and rigor of these data quality checks should align with purpose so that more
stringent checks and supports are in place when states or systems use CLASS for more consequential
decisions. However, Teachstone recommends one or more forms of data quality checks for all CLASS
data collection efforts, as high-quality data is needed to inform improvement efforts. More information on
defining and ensuring reliability can be found in the Technical Summary.
Assigning Observers
One way to ensure that CLASS data is reliable and valid is to randomly assign observers. This approach is
most relevant for large-scale data collection efforts conducted by third parties and is typically not feasible or
recommended for individual program uses of CLASS.
When scheduling randomly, it is essential to randomize across all meaningful units of analysis. For example,
if the data collected is for a program-wide analysis, a random set of observers must be assigned rather than
having the same observer go to each classroom in the program. This helps decrease the possibility of a
single observer inadvertently affecting the accuracy of scores for a whole program. If possible, assign more
than one observer to a site that has more than two or three classrooms.
There are, however, times when observers should be purposefully assigned. When using CLASS at the
program level, strong reasons may exist for coaches or leaders who work with educators to conduct the
observations. Additionally, if not enough observers have the expertise or experience required to observe
in a specific room (for instance, understanding the language of instruction), this warrants more purposeful
assignment.
Double Coding
Double coding (sometimes called double scoring, peer scoring, or shadow scoring) helps ensure that CLASS
coders maintain the integrity of the data collected with CLASS. During double coding, two certified coders
complete observation cycles in the same setting—independently, but simultaneously. They then compare
scores to assess the rate of agreement or the percentage of dimension scores that are within one point
of each other. An agreement of 80% or above is acceptable. For data used for consequential decisions
or research purposes, Teachstone recommends double-coding at least 10% of the observations. Double
coding assesses consistency across data collection, but it does not provide specific information about the
individual observer’s performance. Observers are comparing their scores to one another, not to master-
coded video, as they would in certification or calibration.
Fidelity Checks
Organizations may choose to conduct fidelity checks to ensure that observers are following the protocol
when conducting observations. These consist of an experienced observer watching another observer to
determine whether they follow all steps of the protocol (both general coding procedure and project-specific
elements). If the additional coder is trained to conduct fidelity checks, they can do so in conjunction with
double coding. If an observer does not follow protocol during a fidelity check, they may receive additional
training prior to observing again. Organizations should inform observers of the outcome and provide
feedback to support increased fidelity to the protocol.
Calibration
Teachstone recommends that observers calibrate using their consensus-coded videos. Observers will
watch a video that has been coded by a group of expert coders and will assign scores, which are then
compared to the master codes. At a minimum, observers should calibrate once prior to data collection
and again halfway through data collection. If observers are conducting a large number of observations
or observations will inform high-stakes decisions, more frequent calibration to guard against coder drift is
strongly recommended. In these cases observers should calibrate monthly during data collection windows.
An acceptable agreement rate is 80% or above.
Providing Feedback to Educators

After conducting an observation, the observer or the data platform will calculate scores and create reports
as a part of providing feedback to classrooms or programs. The same care and thought that goes into
planning and collecting CLASS data should also be used in reporting it. CLASS data has a variety of
purposes, including professional development, program evaluation, and accountability systems. Regardless
of the purpose, observers should communicate data in a clear and concise manner that allows the
organization to readily translate scores into actions that lead to improvements in teaching and learning.
Reporting CLASS Data by Audience

Identify the audience and purpose of a report in advance and tailor the report accordingly. For example,
administrators may need more information about the overall program, such as comparisons across
classrooms that inform program-wide accountability or CQI efforts. In contrast, reports for educators focus
on their own classrooms and are generally used for professional development. Regardless of the audience,
those receiving a report must have enough information about the CLASS framework to understand the
results. Furthermore, reports should include key information such as the date of the observation, activities
observed, descriptions of the domains or dimensions for the age group observed, and the results (scores
or ranges) at the appropriate level. Finally, observers should deliver feedback within a few weeks after the
observation to best support buy-in and identify next steps.
This section provides information on how to calculate scores, develop reports, and use data to inform
improvement. It also presents some recommendations for using CLASS for CQI reporting. For more details,
refer to the document Best Practices for Reporting CLASS Data.
Classroom-Level Scores
At the classroom level, educators and the leaders who support them can use their data to guide improvement.
Reports can support individual or team professional development. To calculate observation scores, the
observer should first create an average score for each dimension. To get a dimension-level score across
cycles, add up the dimension scores for each cycle and divide by the number of cycles completed. This will
give the average individual cycle scores for each dimension. Then calculate the scores for each domain by
adding up the dimension-level scores and dividing by the number of dimensions within a given domain. These
results represent the average of each of the corresponding dimension scores. Table 4.2 outlines this process
and provides a sample calculation. The CLASS Score Sheets provide an observation summary page for doing
these calculations.
Note that when arriving at the domain score for Emotional Support, the average score for Negative Climate
is reverse-coded because, unlike the other dimensions, a low-range score represents higher effectiveness.
To reverse the score, subtract the average NC score from 8, as this allows the score to fall within the 1 to 7
range. This value is used when calculating the domain-level score.
Classroom Reports
Classroom-level reports are meant to reflect the experiences of all educators and children in a classroom,
unless otherwise specified in the project-specific protocol (for example, focused on the primary educator or
a cohort of children). Feedback should build on strengths but must also include constructive information on
areas for growth that can inform goal setting and lead to relevant professional development.
Organizations must decide whether to share dimension scores or ranges with educators. Educators who
have a solid understanding of CLASS, including the 1–7 scoring range, are in a better position to see their
classroom’s scores than educators who have less knowledge of, or experience with, the tool. In this case it
may be more beneficial to share ranges instead of scores. An additional reason to share dimension scores is
that educators can see exactly where their setting falls for each CLASS dimension, allowing them to direct
their focus toward improving interactions. Similarly, sharing indicator ranges with educators will also provide
more detailed information and guide decisions about professional development.
Table 4.2. Calculating Classroom-Level Summary Scores
Step How Example

1. Calculate each dimension Total the cycle scores for one PC Cycle 1 = 6
average from the cycle scores. dimension. PC Cycle 2 = 7
PC Cycle 3 = 6
Divide by the number of cycles.
PC Cycle 4 = 5
Repeat for the remaining nine
6 + 7 + 6 + 5 = 24
dimensions.
24 / 4 = 6 (PC average)
2. Calculate the reverse-coded Subtract the NC average from 8. NC average = 2

average for Negative Climate.
8 – 2 = 6 (NC average reverse-
coded)
3. Calculate each domain Total the dimension averages for PC average = 6

average from the dimension one domain. NC average reverse-coded = 6
averages.
Divide by the number of EdS average = 5.5
dimensions in the domain. RCP average = 4
Repeat for the remaining two 6 + 6 + 5.5 + 4 = 21.5

domains. 21.5 / 4 = 5.38 (ES average)
CLASS data is most meaningful when it’s paired with ongoing professional development. Sharing results
provides a starting point for discussion, feedback, and determining next steps. For this reason, Teachstone
recommends that educator reports provide dimension-level feedback that includes specific, descriptive
information about the observations.
Another way to provide relevant feedback to educators is by using the Activity Setting items to identify
patterns in interactions. Reports may note that certain indicators or dimensions score higher during specific
formats or content areas. For example, Positive Climate may be higher during morning meeting than it is
during small-group instruction. Or Concept Development scores may be higher during science activities
than at any other time of the day. Sharing this information with educators can help them determine how to
expand their use of these more effective interactions across activities and content areas. Reports can also
show educators how items captured by the CLASS Environment can support their interactions. For instance,
providing visual charts or schedules for children to use during transitions or to communicate feelings may
improve educators’ interactions within the dimensions of Productivity and Educator Sensitivity, respectively.
If coaches are supporting educators, they should also have access to indicator ranges, dimension-
level scores, domain-level averages, Activity Setting item ratings, and environment scores if applicable.
Observation notes, when available, can add to coaches’ understanding of the interactions. Coaches can
use this information to help educators identify areas of strength and work with them to increase their
intentionality around these types of interactions. Coaches can also support educators in making changes
related to growth areas. If coaches are using informal observations to support professional development,
the reports may consist more of observation notes or video clips for the educators and coaches to discuss
prior to developing written action plans. Information included in an action plan may range from learning
more about a specific dimension or indicator to planning how to expand the use of effective interactions to
more settings or more consistently across children.
Program Reports
At the program level, reports on observation data can also drive improvement by informing accountability
efforts or highlighting trends across an organization. Thus, the focus at the program level often does not
include details such as individualized written feedback for educators. Administrators and stakeholders
should also receive a description of the data collection method, including information about the time
frame for data collection, the sampling system used to identify classrooms, the number of classrooms
observed, the number of cycles required per observation (two, four, or six), and the observers’
credentials.
The adoption of a composite score—derived from averaging across the dimensions or domains into a
singular measure of quality at the program level—helps guide decisions such as funding for programs or
guidance for families selecting a program for their children. Program-level domain scores and total scores
are the types used most by federal and state agencies to guide use of CLASS scores. Both methods,
developed in partnership with Teachstone, have been found to be valid and reliable.
One approach is to calculate program-level domain scores by averaging each domain score from all
classrooms observed within the program. The Office of Head Start uses this method to inform their
grant-recipient monitoring program. By creating program averages at the domain level, OHS and grant
recipients can offer targeted support for staff and to monitor progress. OHS has also used the program-
level scores to establish thresholds for grant recipients and to guide funding decisions.
Another approach establishes a composite CLASS score by 1) averaging dimension scores across all
classrooms observed within a program to create program-level dimension averages, then 2) averaging
the program-level dimension averages into one total CLASS score. This method offers the advantages
of easing communication, especially with families, and integrating a CLASS total score into an overall
quality-rating system.
For example, Louisiana adopted CLASS as their sole measure of quality for infant, toddler, and preschool
programs receiving federal subsidies. Each year they publish a state report card designed to inform
families when choosing early childhood programs. Rather than expecting families to sort through the
domain-level scores, which may be difficult to interpret, they opted to use a total score called the Early
Childhood Performance Rating. This allows families to quickly gain a sense of a program’s quality
(unsatisfactory, approaching proficient, proficient, high proficient, and excellent). Other states, including
Alabama, Michigan, and Florida, have also integrated, or are working to integrate, a total CLASS score
into their QRIS quality ratings.
Regardless of the method used to calculate scores, the program-level reports allow administrators and
stakeholders to understand how a program is doing overall in terms of CLASS scores and to provide
focus for ongoing professional development.
Summary
When creating a CQI system, including clear observation protocols and data quality assurance is crucial
to ensuring the reliability and validity of the data collected. Additionally, whether collected observations
are live or by video, a diverse group of well-trained observers needs to capture interactions in an
effective and equitable manner. Finally, providing reports to inform professional development helps to
turn data into actionable steps for educators and programs to understand and improve interactions and
support child outcomes.
Chapter 5: ®
Promoting Equitable CLASS

Observations
CLASS has become one part of a much larger system that works to define, measure, and improve the quality
of early childhood and educational settings. The ultimate goal is to ensure that every child has access to
the educational experiences they need to thrive. Research using CLASS has demonstrated that too few
children have access to effective interactions from birth to third grade. Children experiencing poverty, and
from historically marginalized groups, are less likely to experience effective interactions due to lack of access
(Vernon-Feagons et al., 2019; Isaacs, 2012). To ensure that every child experiences effective educator-child
interactions across the early years, the system and the elements within it must work to ensure equity in terms
of access, experiences, and outcomes for the children, families, and educators they serve.
The Aspen Institute Education & Society Program and the Council of Chief State School Officers (2017)
define educational equity in the following way:
Educational equity means that every student has access to the educational resources and
rigor they need at the right moment in their education across race, gender, ethnicity, language,
disability, sexual orientation, family background, and/or family income. (p. 3)
Achieving educational equity is a broad endeavor, requiring the coordinated attention and efforts of federal,
state, and local educational agencies that control resources and set policies as well as commitments and
action from the many organizations and individuals working alongside them. Teachstone commits to acting
where the greatest sphere of influence and opportunity for partnering with others exists. These actions
begin to address the much deeper political, economic, and racial inequities that exist in our daily work and
the lives of the children, educators, and other people who we serve.
In this chapter, Teachstone provides guidance on how observers can prepare themselves for observations
across a diversity of settings and contexts. The Observation Field Guide expands on some of these
important points. The guidance in this manual is only one aspect of support that is needed to make the use
of CLASS more equitable and inclusive. One manual cannot possibly capture all the potential scenarios
and specifics about learning settings. Thus, the CLASS Measurement Suite provides additional training and
support to help observers look beyond their individual experiences to create a broader understanding of
what quality interactions can entail. Additional resources are available for observers to better understand
and use the CLASS measure across a variety of settings.
Equitable Observation Practices

Given the diversity of the children and educators across learning settings, it is critical that observers
● Recognize and reduce bias in CLASS observations
● Develop a diverse and inclusive understanding of effective interactions
● Use CLASS with attention to considerations for a diversity of contexts
Recognizing and Reducing Bias in CLASS Observations

All people, and so all CLASS observers, bring their own beliefs, experiences, and expectations about
behaviors and interactions to every setting they visit. These individual lenses, or ways of seeing and
interpreting the world, can create bias in what they notice, what they miss, judgments about the significance
and effectiveness of interactions, and how they assign CLASS scores. To avoid this, observers must
recognize the ways in which their own biases may influence CLASS observations.
To recognize and reduce bias, observers should engage in an iterative and ongoing process:
● Acknowledge bias.
● Learn strategies for reducing bias.
● Apply those strategies during CLASS observations.
Acknowledge bias. Observers must acknowledge and accept that everyone has biases that affect their
thinking and behavior. Biases may stem from social and cultural characteristics (race, ethnicity, gender, age,
disability status, language/accent) or other individual traits (personality, likes and dislikes) of those being
observed or characteristics of the settings in which children are observed (geography, quality of materials,
learning setting size, organizational affiliations, curriculum). Awareness and acceptance of these biases and
the ways they influence observations and scoring are critical first steps in reducing their impact on the CLASS
observation process.
Learn strategies. Strategies for reducing the effects of bias in daily life and in CLASS observations
include diversifying one’s interactions, experiences, and environments; learning from individuals
who offer different perspectives; engaging in practices that broaden one’s mindset; and increasing
bias literacy. Observers should also understand CLASS-specific strategies for reducing bias during
observations, such as reviewing diverse representations of effective interactions across settings and
focusing on the experiences of every child.
Apply strategies. With an understanding of strategies, observers can develop a plan to put these strategies
into action to reduce bias in their daily lives and during CLASS observations.
Observers should engage in an ongoing process of acknowledging that bias is real and pervasive, learning
strategies for recognizing and reducing bias, and applying those strategies to ensure that CLASS scores are
equitable across a diversity of settings.
In the Observation Field Guide, Teachstone provides observers with actionable guidance to reduce the
impact of biases before and during observations as well as during scoring. Teachstone also provides
supplemental training in these areas as well as more extensive opportunities for observers to better
understand strategies for reducing bias in CLASS.
In many circumstances, observers will be able to follow the scoring process and ensure that biases have
not significantly influenced their notes or scores. However, if observers believe that a significant lack of
objectivity is affecting the observation, they should refer to their project-specific protocols.
Developing a Diverse and Inclusive Understanding of Effective Interactions

Expanded descriptions of effectiveness. The CLASS measure describes interactions that support learning
and development for children with diverse backgrounds, experiences, assets, and needs and across a
variety of setting types. However, these interactions may look different across settings, so all observers
need to work to expand their understanding of effective interactions across a diversity of settings.
Teachstone developed CLASS definitions, indicators, and behavioral markers in consultation with experts
with a diversity of cultural backgrounds to represent a range of settings. Teachstone’s Observation
Field Guide, CLASS Observation Training, and supplemental resources offer extensive written and
video examples. Teachstone encourages certified observers to use these resources to refine their
understanding of what effective interactions can look like as they use the CLASS measure. In addition
to the rich library of examples provided, observers must also remember that the behavioral markers and
examples illustrate how to observe CLASS indicators but are not exhaustive. The CLASS Observation
Training helps observers understand how to expand from these behavioral markers and examples in ways
that maintain alignment with the CLASS indicators and dimensions.
The Observation Field Guide also provides more specific considerations for using CLASS across a diversity
of settings, which include children and educators from varied cultural backgrounds, children with linguistic
variation, children with developmental variation, and varied educational settings and program models.
Although a CLASS observation can capture expansive information, observers should remember that one
measure cannot capture all aspects of the learning setting. CLASS was not designed to measure specific
culturally and linguistically responsive teaching practices or the implementation of strategies from an
Individualized Education Program (IEP) for children with disabilities. Nor was it designed to capture the
effectiveness of a specific model or approach. Observers should look for only those behaviors that fit within
the dimensions and indicators as defined in the Observation Field Guide.
The behavioral markers and examples illustrate how to observe

CLASS indicators but are not exhaustive.
Variation in Educational Contexts. Although research suggests that educator-child interactions are
beneficial for all children, nuances to these interactions may vary across individuals, cultural communities,
languages, and settings. For example, there may be differences in how educators and children
demonstrate connection or enjoyment. The more observers understand these variations and the ways
interactions may look within them, the more effectively they can use CLASS across settings and contexts.
CLASS Observation Training, CLASS Observation Supports, and other supplemental resources provide
observers with specific guidance in these areas.
It is beyond the scope of CLASS to explicitly describe the extent of these variations in depth. It is also
beyond any single person to know and understand all variations. Therefore, observers should conduct
observations in settings in which they have experience and expertise. For example, they must, at minimum,
understand the language(s) of instruction.
Educators and children have their own unique abilities, interests, and experiences, which may affect their
interactions. The more observers understand about the individuals they observe and the settings in which
they work, the better able they are to notice and assess the interactions that support children’s learning
and development in the specific context. Although it is possible to conduct CLASS observations knowing
very little about the children, educators, and setting, observers should try to learn about the children and
educators whenever possible. The following details are helpful to know in advance of an observation:
● Number and ages of children, and whether this changes across days of the week or a given day
● Number of children with disabilities or with specific educational plans and goals (such as IEPs), and
their developmental ages and characteristics
● Racial and ethnic composition of children and educators
● Language(s) of instruction, language(s) spoken by the educators and children, and language model
of the setting
● Typical schedule and any activities or times of the day when coding is not appropriate
● Other adults who may be present, such as therapists or family members
This is not a comprehensive list of all the information that would be helpful to know about children,
educators, and settings. Observers should refer to additional training materials and resources for more
detailed information about observing in specific settings. Observers should also obtain additional
experience and training to continue to refine their skills and read through these recommendations, as
relevant, prior to observing and scoring.
Using CLASS with Attention to Considerations for a Diversity of Contexts

The Observation Field Guide provides specific recommendations for using CLASS across a diversity of
settings and contexts. Please refer to the field guide for more detail on each of the following considerations.
Understand Variation
● Learn about variability across children, educators, and settings.
● Check for biases.
Observe Every Child
● Note the experiences of every child.
● Attend to how children respond.
Align with CLASS
● Look for nonverbal interactions and behaviors.
● Align with dimension and indicator definitions.
The CLASS training materials provide specific support and recommendations to enhance observers’ skills
in using CLASS across a diversity of settings. Observers should reference training resources as needed to
support their CLASS observation and scoring as well as to continue to develop their understanding of these
setting variations and any others they may be observing through other professional learning opportunities.
Summary
Increased equity in CLASS observations is a necessary component to create a more equitable educational
landscape and requires that observers be intentional regarding their practices. Key practices include
● Recognizing and reducing bias in CLASS observations
● Developing a diverse and inclusive understanding of effective interactions
● Using CLASS with an attention to considerations for a diversity of contexts
Teachstone provides guidance on how observers can prepare for observations across a diversity of settings
and contexts. The guidance in this manual is one aspect of support needed to make the use of CLASS more
equitable and inclusive. Additional guidance is available in the Observation Field Guide and elsewhere
within the CLASS Measurement Suite.
Appendix: ®
CLASS Technical
Summary
The Classroom Assessment Scoring System (CLASS) is an observational instrument developed to assess
classroom quality across learning contexts. This technical appendix provides an overview of published
studies that demonstrate the quality and accuracy of CLASS to measure the interactions that matter most to
children. There are multiple versions of the CLASS tool that span from measuring interaction in infant care
settings all the way to high school.
Two versions of the CLASS measure are in broad use: a preschool version and a K–3 version. The focus of
this technical appendix is on the Pre-K and K–3 versions. Teachstone provides a more technical summary of
the reliability and validity of CLASS on teachstone.com and updates the information annually.5
The intent of this appendix is to provide information on CLASS’s measurement properties based on
available data and samples and to answer common questions about the measure. CLASS represents one of
the most extensively used observational measures for preschool through elementary learning settings. It is
a well-validated tool providing evidence indicating that learning settings that obtain higher scores on CLASS
serve children who make greater academic and social progress during the school year. CLASS also has
standard training procedures to ensure a high degree of reliability across coders. Ample evidence indicates
that CLASS provides reliable estimates of classroom quality, even when the length of the observation is just
a few hours.
The following sections provide a brief overview of CLASS’s quality and accuracy in measuring interactions.
General Properties of the CLASS Measure

Researchers and educators have gathered much information on the quality and accuracy of the CLASS
measure. Since 2004, CLASS has been used in more than 400 studies in tens of thousands of learning
settings in all 50 states and internationally (for example, Finland, Israel, Kazakhstan, Mexico, Quebec,
Australia, and Ecuador) to support CQI efforts and build evidence of CLASS’s effectiveness. Across these
studies CLASS has consistently measured classroom interactions and has been linked with child outcomes.
CLASS 2nd Edition updates focused on clarifying, expanding, and adding context to the CLASS framework.
However, the measure itself has not undergone any significant changes; Teachstone only updated the
guidance and support for its use. Across the field, implementation of CLASS varies in the number of
observation cycles as well as the observation method (that is, in-person, live-remote, and video). The
following sections feature commonly asked questions related to using CLASS and its accuracy.
What information does CLASS provide about learning settings?

The foundation of CLASS is a framework that assesses classroom interactions that matter most to
children’s development. Structurally, CLASS is divided into three domains—Emotional Support, Classroom
Organization, and Instructional Support—with dimensions and indicators under each. Several studies have
demonstrated that the three domains are related to each other; however, each domain also measures
unique aspects of classroom interactions.
5
ue to COVID-19-related limitations, there has been limited reliability or validation research on CLASS 2nd Edition. However, because the core
D
structure of the CLASS framework remains the same, the updates will likely not undermine existing correlations between CLASS and children’s
learning and development. Based on prior research, these correlations may increase, particularly with the inclusion of Activity Setting items and
greater specificity around how to score across diversity of settings.
Average CLASS Scores
Figure A.1. Average CLASS Scores at the Domain Level
How do the CLASS dimensions relate to one another?

Research has found that CLASS dimensions are moderately to highly correlated with one another. CLASS
uses a theoretical framework suggesting three major domains of classroom process (see Chapter 1). In
addition to the correlations between dimensions, the domains have a strong theoretical basis, and CLASS
domain composites are representative of the constructs they measure.
What are good benchmarks to use for CLASS assessments?

Benchmarks help to establish a goal or thresholds for educators. There are three ways to use comparison
benchmarks for CLASS to support an organization’s goals.
First, research has identified the CLASS scores that are optimal for children to see the greatest benefit.
These scores are 5.0 for Emotional Support, 5.0 for Classroom Organization, and 3.25 for Instructional
Support. These benchmarks can guide and measure progress.
A second way to use benchmarks is by comparing program averages with the Head Start monitoring data
(average scores shared in Figures A.1 and A.2). Data is available from 2016 through 2019 at the OHS’s Early
Childhood Learning & Knowledge Center (ECLKC). The most recent data (2019) can act as a benchmark for
programs. Non-Head Start programs may use the ECLKC data for thresholds; however, programs should
consider the extent to which their program demographics match those of Head Start.
Average CLASS scores from the OHS monitoring data can serve as a guiding reference for programs and
educators to better understand how their interactions compare to other programs, within and outside of
their organization. Averages can also be used to establish educator goals to support the CQI systems.
While there is no single large-scale representative study that establishes average scores on CLASS, studies
occurring at the national and state levels (for example, OHS monitoring data and Louisiana QRIS data)
provide insight into possible expectations within Head Start and diverse publicly funded pre-K programs.
The average domain and dimension scores across these groups show similar patterns (see Figure A1). The
Emotional Support and Classroom Organization domains also have higher averages in comparison to the
Instructional Support domain.
Figure A.2. Average CLASS Scores at the Dimension Level
Louisiana has been using CLASS to support their continuous quality improvement model statewide since
2016, and their average scores on Instructional Support have increased from 3.68 in 2016 (their initial year
of conducting statewide CLASS observations) to 4.02 in 2019.
At the dimension level, the OHS monitoring data offers insight into average scores to guide expectations
and goals. Average scores at the dimension level help to create a more holistic picture of the types and
quality of interactions that are occurring in Head Start settings. The overall averages reflected in Figure A.2
are slightly higher than averages found in early CLASS studies.
The final method used to identify benchmarks is program-specific and works best where multiple years of
CLASS data is available. Comparisons to help establish benchmarks include the differences in scores from
fall to spring and trends in the average across multiple years.
Reliability
Reliability refers to how consistent the CLASS measure is in terms of rating educator-child interactions
across observers, time, and settings. CLASS measures interactions that matter most for children from
birth through high school. It is also an accurate measure of interactions across contexts, such as Head
Start programs, state-based pre-K classrooms, and family childcare settings. Teachstone has established
a rigorous certification and calibration process to ensure high levels of interrater reliability for CLASS
observers. Interrater reliability refers to the level of agreement between multiple observers. This section
provides an overview of research related to the reliability of CLASS.
How do people become reliable users of the CLASS measure?

Reliability is the degree to which an instrument is free from random error associated with the process of
measuring the construct of interest. Research on CLASS and other classroom observation tools indicates
that observers may introduce bias into scores. This may be through their mood, previous classroom
experience, known and unknown biases they bring into the observation, or simply variation in how they
define constructs of the measure.
One way to minimize random errors in CLASS assessments is to provide potential observers with training
and materials so they have a clear and comprehensive understanding of the instrument’s purposes and
procedures. Teachstone has established a rigorous certification and calibration process for CLASS to
reduce the potential errors or biases of individual observers (see Chapter 4).
Teachstone’s reliability tests have produced an average interrater reliability (within one point of “master”
codes) of 87%. By comparison, the standards for interrater reliability for classroom observations typically
range from 70% to 80% agreement.
How much consistency is there across users of the CLASS measure?

Determining consistency across observers involves having two observers use CLASS in the same room
simultaneously, then comparing their scores. The level of consistency within observers is based upon
research across the domains and dimensions. Most studies require observers to reach and maintain
agreement levels of 80% within 1 of a “master” coded video, which generally aligns to the CLASS
certification process.
How stable are CLASS scores?

There are many approaches to thinking about the stability of CLASS scores. Overall, CLASS scores are
fairly consistent across cycles as well as time, as measured by the week and the school year. The sections
that follow share findings for each type of stability.
Across Cycles
Many users of CLASS want to minimize the number of observation cycles to reduce cost. The stability
of CLASS scores across cycles has been assessed in preschool and third-grade classrooms. Findings
suggest that CLASS scores are highly stable across time. Observations for the recommended four cycles
require about two hours to provide an adequate sampling of classroom quality across the three CLASS
domains for one day of instruction (Early et al., 2005). This stability can also be found at the CLASS
dimension level, with the relationship between cycle dimension scores and the final score ranging from
moderate to high. In addition, a group of researchers conducted multiple CLASS cycles over two to three
days and found that four cycles provided a representative sample of classroom quality on which to base
final CLASS scores (NICHD ECCRN, 2002, 2005a).
Another way of examining stability is by assessing the degree of internal consistency of CLASS scores
across two, three, and four cycles. Mirroring the previously mentioned results, these results suggest a
high degree of internal consistency for the CLASS scores across cycles (Early et al., 2005; NICHD ECCRN,
2002, 2005a). Internal consistency is somewhat higher among the Emotional Support dimensions than
among the Classroom Organization or Instructional Support dimensions (Li et al., 2020).
Across Days in the Week

Another question that concerns observers is the extent to which one day is representative of general
classroom practices. One study observed preschool learning settings on two, typically consecutive,
days in the spring of the school year. Their results suggested a high degree of stability, with correlations
between the two days ranging from .73 for Productivity to .85 for Teacher Sensitivity (Early et al., 2005).
Small but significant mean changes occurred across several of the dimensions, with a general trend
toward lower-quality scores on the second day. Given that there is no reason to expect a systematic
difference in quality across two consecutive days, these small changes may be due to observer bias in
which scores become slightly lower over time. Again, however, although these differences are statistically
significant, they are relatively small effects; correlations between the two days are high.
In another study across more than 200 cases, the same classroom/educator was observed on more than
one occasion (NICHD ECCRN, 2002, 2005a). Researchers found that CLASS ratings across days and
groups of children were highly correlated: all exceeded .65. These analyses indicate that 7-point ratings
of the learning setting and educator behavior are highly stable and are not highly dependent on the
occasion or reason for observing.
Across the School Year

Scores on CLASS are also relatively stable from fall to spring, at least across a large number of preschool
programs (Pianta et al., 2007). Only small changes occurred in mean scores, and correlations between
fall and spring range from .64 for Behavior Management to .25 for Quality of Feedback. In general, there
appears to be lower levels of stability in the Instructional Support dimension than in other dimensions,
with Concept Development and Quality of Feedback decreasing from fall to spring. Overall, findings
suggest that CLASS may be sensitive to the fluctuations in quality that are often described around the
holidays and toward the end of the school year.
Why do CLASS scores change based on the content being taught and throughout the
same day?
It is true that CLASS scores may shift based on the time of day (that is, start of day and transitions) and
the activity setting (for example, literacy or math instruction). But these scores are due to the typical shifts
in educator behaviors linked to interactions rather than an inconsistency in the CLASS scores themselves.
For example, an educator may focus on more managerial interactions (for example, putting items in
cubbies, preparing for the lesson, taking attendance) at the beginning of the day and during transitions,
which could lead to lower scores for the Instructional Support domain.
Scores may also vary depending on the activity setting. For instance, during math lessons, educators
may be more rigid and focused on rote learning in comparison to a literacy activity where educators
invite children to contribute a story and guide the activity. A study conducted by Curby and colleagues
(2011) demonstrated that CLASS scores are most stable throughout the day in third- through fifth-grade
classrooms for the Emotional Support domain and dimensions. The study also showed that CLASS
scores are least stable within the Instructional Support domain and dimensions. Their findings indicated
that interactions at the start of the day and during transitions were associated with lower scores on
the Instructional Support dimensions. The findings by activity setting were consistent with those. The
expansion of CLASS to include additional detail related to activity setting will allow a closer look at these
aspects that are linked to interactions and support CQI.
How reliable are video observations using CLASS?

CLASS was originally designed for use in live observations. However, a shift has occurred over time to
include video and remote-live options as observation methods. Studies have compared CLASS scores
assigned through simultaneous live and video observations (Curby et al., 2016). While slight variations occur
across the studies, the overall findings suggest there is little to no difference in the CLASS scores assigned
through live or video observations on the Pre-K CLASS measure. A large-scale study using the Secondary
CLASS tool found that live observations had higher levels of reliability (Casabianca et al. 2013) over video
observations. However, using video is still recommended since it offers a tenable approach for scaling
observations.
Another approach in the field is the use of remote-live observations. This method uses a camera controlled
by an individual in another location. The reliability of remote-live observations using the CLASS measure as
compared to live and video observations is still under investigation.
Validity
Validity focuses on the extent to which CLASS accurately measures educator-child interactions. The
accuracy of the CLASS measure allows for inferences about the types of interactions that are occurring in
learning settings and their potential impact on children’s social, emotional, and academic development.
Does CLASS measure constructs of importance in learning settings (face and construct
validity)?
CLASS was developed based on an extensive literature review of classroom practices related to
children’s social and academic development in schools. The dimensions were derived from a review
of constructs assessed in classroom observation instruments used in childcare and elementary school
research, literature on effective teaching practices, feedback from focus groups, and data from extensive
piloting. Throughout this process, numerous experts in classroom quality and teaching effectiveness
agreed that CLASS measures aspects that are important in determining children’s performance,
suggesting considerable face validity.
How does CLASS relate to other measures of classroom quality and associated
constructs (criterion validity)?
Criterion validity assesses the extent to which a measure is empirically associated with other measures
of similar constructs. As demonstrated in research with CLASS, settings with higher Emotional Support,
Classroom Organization, and Instructional Support received higher scores on two factors of the Early
Childhood Environment Rating Scale, Revised Edition (ECERS-R), which is another commonly used
measure of quality in early childhood learning settings (Harms et al., 2014). The ECERS-R Interaction factor
describes the extent to which learning settings promote educator-child interactions, encourage children
to communicate and use language, and provide effective discipline. The Space and Furnishings factor
assesses the availability of furnishings and materials. Consistent with the CLASS focus on classroom
interactions, it is not surprising that CLASS had stronger associations with the Interaction factor.
The Emerging Academics Snapshot (Ritchie et al., 2001) is a time-sampling method used to assess the
percentage of time spent on various activities. Within the Multistate Study of state-funded pre-K programs,
classrooms with higher CLASS scores spent more time on literacy and language than on math activities. The
somewhat lower correlations between the CLASS Instructional Support domain and time spent in literacy
and math are not surprising, given that this domain focuses on quality, not quantity, of instruction. Children
in learning settings with higher CLASS scores spent more time engaged in elaborate interactions with
adults for significantly more time.
The Measures of Effective Teaching (MET) study provides insight into the correlation between the
CLASS and other measures used in upper elementary and secondary classrooms (Kane et al., 2013). The
Framework for Teaching (FFT; Danielson, 2014) is a general pedagogical instrument like CLASS. Other
measures used by the MET study focused on the quality of math and language arts instruction. All the
measures had high correlations with CLASS; however, the strongest correlation was with the FFT. This
suggests that CLASS captures similar constructs to other measures of instructional quality.
How does CLASS relate to children’s academic and social development (predictive
validity)?
CLASS assesses classroom-level interactions that are directly associated with children’s performance.
A growing body of research has demonstrated the link between child outcomes across a diverse
set of programs and children and interactions, as measured by CLASS. However, findings about
how well CLASS predicts child outcomes have been mixed. Qualities of educator-child interactions,
including sensitivity to individual needs, support for child engagement, and stimulation of conceptual
understanding promote child outcomes across pre-K–12 (Allen et al., 2011; Hamre et al., 2013; Pianta et
al., 2010). Interactions that foster conceptual understanding and provide feedback extend children’s skills
and predict gains in learning in the later elementary grades (Guo et al., 2012; Reyes et al., 2012). High-
quality interactions with educators predict children’s academic attitudes, satisfaction with school, and
increased motivation (Klem & Connell, 2004; Martin & Rimm-Kaufman, 2015). Children who display some
form of risk (for example, low income, limited English skills, problems with self-regulation) differentially
benefit from effective educator-child interactions (Griggs et al., 2016; Hamre & Pianta, 2005; Shin & Ryan,
2017), particularly when present across multiple years (Pianta et al., 2008; Vernon-Feagans et al., 2019).
As noted through these studies, classroom interactions are a pivotal factor in children’s development.
The sections that follow share research related to the ability to use CLASS to predict children’s social-
emotional and academic outcomes.
An expanding body of research demonstrates the relationship between CLASS and children’s outcomes from
birth through adolescence. Within higher-quality pre-K and kindergarten learning settings, as measured by
CLASS, children demonstrate higher school readiness skills such as language, literacy, math, working memory,
and inhibitory control (Carr et al., 2019; Hamre et al., 2014; Vitiello et al., 2018). Studies focused on infants and
toddlers have found a relationship between CLASS and outcomes such as child well-being (Bratsch-Hines et
al., 2020). For instance, higher emotional support scores relate to improved child language development. The
relationship between CLASS scores and student outcomes has also been established for middle and high
school students (Allen et al., 2013). In addition, the MET study found that educators who demonstrated the
types of practices emphasized in CLASS had higher value-added scores than educators who did not.
More nuanced relationships can be found at the dimension level, with social and emotional support being
predictive of children’s social skills, classroom engagement, and behavior (for example, Broekhuizen et al.,
2017; Choi et al., 2016; Pakarinen et al., 2020). Children in settings with higher Classroom Organization show
increased social adaptation, less disruptive play, and increased inhibitory control (Besnard & Letarte, 2017;
Bulotsky-Shearer et al., 2020; Moiduddin et al., 2012). Finally, children in settings with higher Instructional
Support scores show a greater ability to regulate their behavior, engage in higher-order thinking skills,
and demonstrate higher executive functioning skills and higher levels of school readiness skills (for
example, Alamos & Williford, 2020; Cadima et al., 2016; Cash et al., 2019). Within toddler learning settings,
higher levels of emotional and behavioral support were more strongly associated with positive social and
emotional outcomes as well as improved language development (Aikens et al., 2015).
The CLASS measure has also demonstrated predictive validity with children who are DLLs. Studies have
demonstrated a link between the Emotional Support domain and increased social competence among
DLLs (Downer et al., 2011) as well as higher reading and math scores (Burchinal et al., 2012). The FACES
study, based on a nationally representative sample of Head Start programs, demonstrated that Instructional
Support was positively correlated with children’s development of both Spanish and English vocabulary
(Hindman & Wasik, 2013).
Evidence has also demonstrated predictive validity of CLASS across cultural communities, including Latin
America, Asia, and Europe. A multinational synthesis of studies across 16 countries demonstrated that
CLASS could be reliably observed and had significantly beneficial impacts on children’s learning (Hofkens et
al., in press). The results were like those found within the United States.
While there is a clear pathway for the impact of interactions on children’s development and evidence of
the predictive validity of CLASS from close to 200 studies, it is worth noting that the effect sizes are small
to moderate in most cases. In addition, some studies have found no relationship between CLASS and
child outcomes (Guererro-Rosada et al., 2021). The variation in effect sizes can be attributed to the timing
or frequency of CLASS observations, the sensitivity and alignment of the measures used to assess child
outcomes, and other contextual factors including the curriculum used.
Summary
Ratings from the CLASS observation measure provide important descriptive information about the current
status of early childhood and elementary learning settings and also have a predictive value in academic
and social outcomes for children. As such, CLASS can be a useful tool for researchers, administrators, and
policymakers wanting a standardized measure of classroom processes that are empirically linked to important
child outcomes. See teachstone.com for more detailed information and updates.
References
Aikens, N., Bush, C., Gleason, P., Malone, L., & Tarullo, L. (2016). Tracking Quality in Head Start Classrooms: FACES
2006 to FACES 2014 Technical Report. OPRE Report 2016-95. Office of Planning, Research and Evaluation,
Administration for Children and Families, U.S. Department of Health and Human Services.
Aikens, N., Xue, Y., Bandel, E., Caronongan, P., Vogel, C. A., & Boller, K. (2015). Early Head Start home visits
and classrooms: Stability, predictors, and thresholds of quality, Baby FACES 2009. OPRE Report 2015–34.
Washington, DC: Office of Planning, Research and Evaluation, Administration for Children and Families, U.S.
Department of Health and Human Services.
Alamos, P., & Williford, A. P. (2020). Exploring dyadic teacher-child interactions, emotional security, and task
engagement in preschool children displaying externalizing behaviors. Social Development, 29(1), 338–355.
Allen, J. P., Hafen, C. A., Gregory, A. C., Mikami, A. Y., & Pianta, R. C. (2015). Enhancing secondary school instruction
and student achievement: Replication and extension of the My Teaching Partner-Secondary intervention. Journal
of Research on Educational Effectiveness, 8(4), 475–489.
Allen, J., Gregory, A., Mikami, A., Lun J., Hamre, B., Pianta, R. (2013). Observations of effective teacher-student
interactions in secondary school classrooms: Predicting student achievement with the classroom assessment
scoring system-secondary. School Psychology Review, 42(1), 76–98.
Allen, J. P., Pianta, R. C., Gregory, A., Mikami, A. Y., & Lun, J. (2011). An interaction-based approach to enhancing
secondary school instruction and student achievement. Science, 333(6045), 1034–1037.
Araujo, M. C., Carneiro, P., Cruz-Aguayo, Y., & Schady, N. (2016). Teacher quality and learning outcomes in
kindergarten. Quarterly Journal of Economics, 131(3), 1415–1453.
Bassok. D., & Markowitz, A. J. (2020). The value of systemwide, high-quality data in early childhood education. https://
www.brookings.edu/blog/brown-center-chalkboard/2020/02/20/the-value-of-systemwide-high-quality-data-in-
early-childhood-education/
Besnard, T., & Letarte, M-J. (2017). Effect of male and female early childhood education teachers’ educational
practices on children’s social adaptation. Journal of Research in Childhood Education, 31(3), 453–464.
Bratsch-Hines, M. E., Carr, R., Zgourou, E., Vernon-Feagans, L., & Willoughby, M. (2020). Infant and toddler child-care
quality and stability in relation to proximal and distal academic and social outcomes. Child Development, 91(6),
1854–1864.
Broekhuizen, M. L., Slot, P. L., van Aken, M. A. G., & Dubas, J. S. (2017). Teachers’ emotional and behavioral support
and preschoolers’ self-regulation: Relations with social and emotional skills during play. Early Education and
Development, 28(2), 135–153.
Broekhuizen, M. L., Mokrova, I. L., Burchinal, M. R., Garrett-Peters, P. T., & Family Life Project Key Investigators. (2016).
Classroom quality at pre-kindergarten and kindergarten and children’s social skills and behavior problems. Early
Childhood Research Quarterly, 36, 212–222.
Brunsek, A., Perlman, M., Falenchuk, O., McMullen, E., Fletcher, B., & Shah, P. S. (2017) The relationship between the
Early Childhood Environment Rating Scale and its revised form and child outcomes: A systematic review and
meta-analysis. PLoS ONE, 12(6): Article e0178512. https://doi.org/10.1371/journal.pone.0178512
Bulotsky-Shearer, R. J., Fernandez, V. A., Bichay-Awadalla, K., Bailey, J., Futterer, J., & Qi, C. H. (2020). Teacher-child
interaction quality moderates social risks associated with problem behavior in preschool classroom contexts.
Journal of Applied Developmental Psychology, 67, Article 101103. https://doi.org/10.1016/j.appdev.2019.101103
Burchinal, M. (2018). Measuring early care and education quality. Child Development Perspectives, 12(1), 3–9.
https://doi.org/10.1111/cdep.12260
Burchinal, M., Field, S., López, M. L., Howes, C., & Pianta, R. (2012). Instruction in Spanish in pre-kindergarten
classrooms and child outcomes for English language learners. Early Childhood Research Quarterly, 27(2), 188–197.
Burchinal., M., Soliday-Hong, S. L., Sabol, T. J., Forestieri, N., Peisner-Feinberg, E., Tarullo, L., & Zaslow, M. (2016).
Quality rating and improvement systems: Secondary data analyses of psychometric properties of scale
development. U.S. Department of Health and Human Services.
Cabell, S. Q., DeCoster, J., LoCasale-Crouch, J., Hamre, B. K., & Pianta, R. C. (2013). Variation in the effectiveness of
instructional interactions across preschool classroom settings and learning activities. Early Childhood Research
Quarterly, 28(4), 820–830. https://doi.org/10.1016/j.ecresq.2013.07.007
Cadima, J., Leal, T., & Burchinal, M. (2010). The quality of teacher-student interactions: Association with first graders’
academic and behavior outcomes. Journal of School Psychology, 48(6), 457–482.
Cadima, J., Verschueren, K., Leal, T., Guedes, C. (2016). Classrooms interactions, dyadic teacher-child relationships,
and self-regulation in socially disadvantaged young children. Journal of Abnormal Child Psychology, 44(1), 7–17.
Carneiro, P., Cruz-Aguayo, Y., & Schady, N. (2019). Experimental estimates of education production functions: Sensitive
periods and dynamic complementarity. Institute for Fiscal Studies, University College London.
Carr, R. C., Mokrova, I. L., Vernon-Feagans, L., & Burchinal, M. R. (2019). Cumulative classroom quality during
pre-kindergarten and kindergarten and children’s language, literacy, and mathematics skills. Early Childhood
Research Quarterly, 47, 218–228.
Casabianca, J. M., McCaffrey, D. F., Gitomer, D. H., Bell, C. A., Hamre, B. K., & Pianta, R. C. (2013). Effect of observation
mode on measures of secondary mathematics teaching. Educational and Psychological Measurement, 73(5), 757-
783.
Cash, A. H., Ansari, A., Grimm, K. J., & Pianta, R. C. (2019). Power of two: The impact of 2 years of high quality teacher
child interactions. Early Education and Development, 30(1),1–22.
Choi, J. Y., Castle, S., Williamson, A. C., Young, E., Worley, L., Long, M., & Horm, D. M. (2016). Teacher-child interactions
and the development of executive function in preschool-age children attending Head Start. Early Education and
Development, 27(6), 751–769.
Clements, D. H., & Sarama, J. (2014). COEMET: The Classroom Observation of Early Mathematics Environment and
Teaching instrument. University of Denver.
Conroy, M. A., Sutherland, K. S., Algina, J., Ladwig, C. N., Werch, B. L., Martinez, J., Jessee, G., & Gyure, M. (2019).
Outcomes of the BEST in CLASS intervention on teachers’ use of effective practices, self-efficacy, and classroom
quality. School Psychology Review, 48(1), 31–45.
Curby, T. W., Brock, L. K., & Hamre, B. K. (2013). Teachers’ emotional support consistency predicts children’s
achievement gains and social skills. Early Education and Development, 24(3), 292–309.
Curby, T. W., Johnson, P., Mashburn, A. J., & Carlis, L. (2016). Live versus video observations: Comparing the reliability
and validity of two methods of assessing classroom quality. Journal of Psychoeducational Assessment, 34(8),
765–781.
Curby, T. W., Stuhlman, M., Grimm, K., Mashburn, A., Chomat-Mooney, L., Downer, J. T., Hamre, B. K., & Pianta, R. C.
(2011). Within-day variability in the quality of classroom interactions during third and fifth-grade. The Elementary
School Journal, 112, 16–37.
Curenton, S. M., Iruka, I. U., Humphries, M., Jensen, B., Durden, T. Rochester, S. E., Sims, J., W, J. V., & Kinzie, M. B.
(2019). Validity for the Assessing Classroom Sociocultural Equity Scale (ACSES) in early childhood classrooms.
Early Education and Development, 31(2), 284–303.
Danielson, C. (2014). The framework for teaching evaluation instrument. Danielson Group. Retrieved January 6, 2015,
from http://www.danielsongroup.org.
Denham, S. A., Bassett, H. H., & Zinsser, K. (2012) Computerizing social-emotional assessment for school readiness:
First steps toward an assessment battery for early childhood settings, Journal of Applied Research on Children,
3(2), Article 3.
Domínguez, X., Vitiello, V. E., Fuccillo, J. M., Greenfield, D. B., & Bulotsky-Shearer, R. J. (2011). The role of context
in preschool learning: A multilevel examination of the contribution of context-specific problem behaviors and
classroom process quality to low-income children’s approaches to learning. Journal of School Psychology, 49(2),
175–195.
Downer, J. T., Booren, L. M., Hamre, B., Pianta, R. C., & Williford, A. (2011). The Individualized Classroom Assessment
Scoring (inCLASS). Unpublished technical manual, Curry School of Education, University of Virginia.
Downer, J. T., Kraft-Sayre, M. E., & Pianta, R. C. (2009). Ongoing, web-mediated professional development focused on
teacher–child interactions: Early childhood educators’ usage rates and self-reported satisfaction. Early Education
and Development, 20(2), 321–345.
Downer, J. T., López, M. L., Grimm, K. J., Hamagami, A., Pianta, R. C., & Howes, C. (2012). Observations of teacher-child
interactions in classrooms serving Latinos and dual language learners: Applicability of the Classroom Assessment
Scoring System in diverse settings. Early Childhood Research Quarterly, 27(1), 21–32.
Early, D., Barbarin, O., Bryant, D., Burchinal, M., Chang, F., Clifford, R., Crawford, G., Weaver, W., Howes, C., Ritchie,
S., Kraft-Sayre, M., Pianta, R., & Barnett, W. S. (2005). Prekindergarten in eleven states: NCEDL’s multi-state study
of prekindergarten and study of state-wide early education programs. Retrieved December 1, 2005, from http://
www.fpg.unc.edu/NCEDL/pdfs/SWEEP _MS_summary_final.pdf
Early, D. M., LaForett, D. R., & Kraus, S. (2017). Evaluation findings from Year 2 of Georgia’s CLASS-Related
Professional Development Evaluation Project. University of North Carolina at Chapel Hill, FPG Child Development
Institute.
Early, D. M., Maxwell, K. L., Skinner, D., Kraus, S., Hume, K., & Pan, Y. (2014). Georgia’s Pre-K Professional Development
Evaluation: Final report. University of North Carolina at Chapel Hill.
Farley, K. S., Piasta, S., Dogucu, M., & O’Connell, A. (2017). Assessing and predicting small-group literacy instruction in
early childhood classrooms. Early Education and Development, 28(4), 488–505.
https://doi.org/10.1080/10409289.2016.1250549
Finders, J. K., Budrevich, A., Duncan, R., Purpura, D. J., Elicker, J., & Schmitt, S. (2021). Variability in preschool CLASS
scores and children’s school readiness. AERA Open, 7(1), 1–19. https://doi.org/10.1177/23328584211038938
Florida Department of Education. (n.d.) Early Learning Performance Funding Project. http://www.floridaearlylearning.
com/statewide-initiatives/early-learning-performance-funding-project
Griggs, M. S., Mikami, A. Y., & Rimm-Kaufman, S. E. (2016). Classroom quality and student behavior trajectories in
elementary school. Psychology in the Schools, 53(7), 690–704.
Guerrero-Rosada, P., Weiland, C., McCormick, M., Hsueh, J., Sachs, J., Snow, C., & Maier, M. (2021). Null relations
between CLASS scores and gains in children’s language, math, and executive function skills: A replication and
extension study. Early Childhood Research Quarterly, 54, 1–12.
Guo, Y., Connor, C. M., Yang, Y., Roehrig, A. D., & Morrison, F. J. (2012). The effects of teacher qualification, teacher
self-efficacy, and classroom practices on fifth graders’ literacy outcomes. The Elementary School Journal, 113(1),
3–24.
Guo, Y., Piasta, S. B., Justice, L. M., & Kaderavek, J. N. (2010). Relations among preschool teachers’ self-efficacy,
classroom quality, and children’s language and literacy gains. Teaching and Teacher Education, 26(4), 1094–1103.
https://doi.org/10.1016/j.tate.2009.11.005
Hafen, C. A., Hamre, B. K., Allen, J. P., Bell, C. A., Gitomer, D. H., & Pianta, R. C. (2015). Teaching through interactions
in secondary school classrooms: Revisiting the factor structure and practical application of the Classroom
Assessment Scoring System–Secondary. Journal of Early Adolescence, 35, 651–680.
Hamre, B., Hatfield, B., Pianta, R., & Jamil, F. (2014). Evidence for general and domain-specific elements of teacher-
child interactions: Associations with preschool children’s development. Child Development, 85(3), 1257–1274.
Hamre, B. K. (2014). Teachers’ daily interactions with children: An essential ingredient in effective early childhood
programs. Child Development Perspectives, 8(4), 223–230.
Hamre, B. K., Justice, L. M., Pianta, R. C., Kilday, C., Sweeney, B., Downer, J. T., & Leach, A. (2010). Implementation
fidelity of MyTeachingPartner literacy and language activities: Association with preschoolers’ language and
literacy growth. Early Childhood Research Quarterly, 25, 329–347.
Hamre, B. K., & Pianta, R. C. (2005). Can instructional and emotional support in the first grade classroom make a
difference for children at risk of school failure? Child Development, 76(5), 949–967.
Hamre, B. K., & Pianta, R. C. (2007). Learning opportunities in preschool and early elementary classrooms. In
R. C. Pianta, M. J. Cox, & K. L. Snow (Eds.), School readiness and the transition to kindergarten in the era of
accountability (pp. 49–83). Paul H Brookes Publishing.
Hamre, B. K., Pianta, R. C., Burchinal, M., Field, S., LoCasale-Crouch, J., Downer, J. T., Howes, C., LaParo,
K., & Scott-Little, C. (2012a). A course on effective teacher-child interactions: Effects on teacher beliefs,
knowledge, and observed practice. American Education Research Journal, 49(1), 88–123. http://dx.doi.
org/10.3102/0002831211434596
Hamre, B. K., Pianta, R. C., Downer, J. T., DeCoster, J., Mashburn, A. J., Jones, S. M., Brown, J. L., Cappella, E.,
Atkins, M., Rivers, S. E., Brackett, M. A., & Hamagami, A. (2013). Teaching through interactions: Testing a
developmental framework of teacher effectiveness in over 4,000 classrooms. The Elementary School Journal,
113(4), 461–487.
Hamre, B. K., Pianta, R. C., Mashburn, A. J., & Downer, J. T. (2012b). Promoting young children’s social competence
through the preschool PATHS curriculum and MyTeachingPartner professional development resources. Early
Education and Development, 23(6), 809–832.
Han, J., Schlieber, M., & Gregory, B. (2017). Associations of home and classroom environments with Head Start
children’s code-related and oral language skills. Journal of Education for Students Placed at Risk, 22(4), 200–219.
Harms, T., Clifford, R. M., & Cryer, D. (2014). Early Childhood Environment Rating Scale, third edition (ECERS-3). https://
ers.fpg.unc.edu/scales-early-childhood-environment-rating-scale-third-edition
Hatfield, B. E., Finders, J. K., Zandbergen, D. L., & Lewis, H. (2021). Associations between consistent and high-quality
teacher-child interactions and preschool children’s self-regulation and activity in the stress response system. Early
Education and Development. https://doi.org/10.1080/10409289.2021.1961198
Hatfield, B. E., Hestenes, L. L., Kintner-Duffy, V. L., & O’Brien, M. (2013). Classroom emotional support predicts
differences in preschool children’s cortisol and Alpha-amylase levels. Early Childhood Research Quarterly, 28(2),
347–356.
Hindman, A. H., & Wasik, B. A. (2013). Vocabulary learning in Head Start: Nature and extent of classroom instruction
and its contributions to children’s learning. Journal of School Psychology, 51(3), 387–405. https://doi.org/10.1016/j.
jsp.2013.01.001
Hindman, A. H., & Wasik, B. A. (2015). Building vocabulary in two languages: An examination of Spanish-speaking Dual
Language Learners in Head Start. Early Childhood Research Quarterly, 31, 19–33.
Hirshberg, M. J., Flook, L., Enright, R. D., & Davidson, R. J. (2020). Integrating mindfulness and connection practices
into preservice teacher education improves classroom practices. Learning and Instruction, 66, Article 101298.
https://doi.org/10.1016/j.learninstruc.2019.101298
Hofkens, T., Pianta, R. C., & Hamre, B. K. (in press). Teacher-student interactions: Theory, measurement, and evidence
for universal properties that support students’ learning across countries and cultures. In R. Maulana (Ed.), Effective
teaching around the world: Theoretical, empirical, methodological and practical insights. Springer.
Howes, C., Burchinal, M., Pianta, R. C., Bryant, D., Early, D., Clifford, R., & Barbarin, O. (2008). Ready to learn? Children’s
pre-academic achievement in pre-kindergarten programs. Early Childhood Research Quarterly, 23, 27–50.
Hu, B. Y., Fan, X., Gu, C., & Yang, N. (2016). Applicability of the Classroom Assessment Scoring System in Chinese
preschools based on psychometric evidence. Early Education and Development, 27(5), 714–734.
https://doi.org/10.1080/10409289.2016.1113069
Hu, B. Y., Fan, X., Wu, Z., LoCasale-Crouch, J., Yang, N., & Zhang, J. (2017). Teacher-child interactions and children’s
cognitive and social skills in Chinese preschool classrooms. Children and Youth Services Review, 79, 78–86.
https://doi.org/10.1016/j.ecresq.2019.10.003
Hu, B. Y., Huang, P., Wang, S. & Curby, T. (2021). Teachers’ beliefs about children and children’s literacy development:
The mediating role of responsive teaching. Journal of Early Childhood Literacy.
https://doi.org/10.1177/14687984211037004
Isaacs, J. B. (2012). Starting school at a disadvantage: The school readiness of poor children. The Social Genome
Project. Center on Children and Families at Brookings.
Johnson, A. D., Markowitz, A. J., Hill, C. J., & Phillips, D. A. (2016). Variation in impacts of Tulsa pre-K on cognitive
development in kindergarten: The role of instructional support. Developmental Psychology, 52(12), 2145–2158.
Johnson, S. R., Seidenfeld, A. M., Izard, C. E., & Kobak, R. (2013). Can classroom emotional support enhance prosocial
development among children with depressed caregivers? Early Childhood Research Quarterly, 28, 282–290.
Kane, T. J., McCaffrey, D. F., Miller, T., & Staiger, D. O. (2013). Have we identified effective teachers? Validating
measures of effective teaching using random assignment. [Research Paper]. MET Project. Bill & Melinda Gates
Foundation.
Kane, T. J., & Staiger, D. O. (2012). Gathering feedback for teaching: Combining high-quality observations with student
surveys and achievement gains. [Research Paper]. MET Project. Bill and Melinda Gates Foundation.
Klem, A. M., & Connell, J. P. (2004). Relationships matter: Linking teacher support to student engagement and
achievement. Journal of School Health, 74(7), 262–273.
Lee, P., & Bierman, K.L. (2016). Profiles of kindergarten classroom and elementary school contexts: Associations with
the first-grade outcomes of children transitioning from Head Start. The Elementary School Journal, 117(1), 119–142.
Leyva, D., Weiland, C., Barata, M., Yoshikawa, H., Snow, C., Treviño, E., & Rolla, A. (2015). Teacher–child interactions
in Chile and their associations with prekindergarten outcomes. Child Development, 86(3), 781–799. https://doi.
org/10.1111/cdev.12342
Li, H., Liu, J., & Hunter, C. V. (2020). A meta-analysis of the factor structure of the Classroom Assessment Scoring
System (CLASS). The Journal of Experimental Education, 88(2), 265–287.
Limlingan, M. C., McWayne, C. M., Sanders, E. A. & Lopez, M. L. (2019). Classroom language contexts as predictors
of Latinx preschool dual language learners’ school readiness. American Educational Research Journal, 57(1),
339–370. https://doi.org/10.3102/0002831219855694
López, F. A. (2012). Moderators of language acquisition models and reading achievement for English Language
Learners: The role of emotional warmth and instructional support. Teachers College Record, 114(8), 1–30.
https://doi.org/10.1177/016146811211400803
Maier, M. F., Vitiello, V. E., & Greenfield, D. B. (2012). A multilevel model of child- and classroom-level psychosocial
factors that support language and literacy resilience of children in Head Start. Early Childhood Research
Quarterly, 27, 104–114.
Markowitz, A. J., Bassok, D., & Player, D. (2020). Simplifying quality rating systems in early childhood education.
Children and Youth Services Review, 112, Article 104947. https://doi.org/10.1016/j.childyouth.2020.104947
Martin, D. P., & Rimm-Kaufman, S. E. (2015). Do student self-efficacy and teacher-student interaction quality contribute
to emotional and social engagement in fifth grade math? Journal of School Psychology, 53(5), 359–373.
Mashburn, A. J., Downer, J. T., Rivers, S. E., Brackett, M. A., & Martinez, A. (2014). Improving the power of an efficacy
study of a social and emotional learning program: Application of generalizability theory to the measurement of
classroom-level outcomes. Prevention Science, 15(2), 146–155.
Mashburn, A. J., Pianta, R. C., Hamre, B. K., Downer, J. T., Barbarin, O. A., Bryant, D., Burchinal, M., Early, D. M., &
Howes, C. (2008). Measures of classroom quality in pre-kindergarten and children’s development of academic,
language, and social skills. Child Development, 79(3), 732–749.
McDoniel, M. E., Townley-Flores. C., Sulik, M., & Obradović, J. (2022) Widely used measures of classroom quality are
largely unrelated to preschool skill development. Early Childhood Research Quarterly, 59,243–253.
Moiduddin, E., Aikens, N. Tarullo, L., West, J., & Xue, Y., (2012). Child outcomes and classroom quality in FACES 2009.
(OPRE report 2013–37a). Office of Planning, Research and Evaluation, Administration for Children and Families,
U.S. Department of Health and Human Services.
Neugebauer, S., Sandilos, L., Coyne, M.,McCoach, D. B., & Ware, S. (2020) Highly potent and vastly conditional
instructional practices: Variations in use and utility of language interactions for kindergarten. Early Education and
Development, 31(4), 541–560. https://doi.org/10.1080/10409289.2019.1686928
NICHD Early Child Care Research Network. (2002). The relation of global first grade classroom environment to structural
classroom features, and teacher and student behaviors. The Elementary School Journal, 102(5), 367–387.
NICHD Early Child Care Research Network. (2003). Social functioning in first grade: Prediction from home, child care
and concurrent school experience. Child Development, 74, 1639–1662.
NICHD Early Child Care Research Network. (2005a). A day in third grade: A large-scale study of classroom quality and
teacher and student behavior. The Elementary School Journal, 105(3), 305–323.
NICHD Early Child Care Research Network. (2005b). Predicting individual differences in attention, memory, and
planning in first graders from experiences at home, child care, and school. Developmental Psychology, 41(1),
99–114.
Niklas, F. & Tayler, C. (2018). Room quality and composition matters: Children’s verbal and numeracy abilities in
Australian early childhood settings. Learning and Instruction, 54, 114–214.
Nores, M., Friedman-Krauss, A., & Figueras-Daniel, A. (2022). Activity settings, content, and pedagogical strategies in
preschool classrooms: Do these influence the interactions we observe? Early Childhood Research Quarterly, 58,
264–277.
Office of Head Start. (2018). Designation Renewal System: DRS by the numbers. Office of Head Start, Administration
for Children and Families, U.S. Department of Health and Human Services. https://eclkc.ohs.acf.hhs.gov/sites/
default/files/pdf/drs-by-the-numbers-report.pdf
Oliva-Olson, C., Espinosa, L. M., Zepeda, M., Durán, L., Fernandez, V., & Arambula-Gonzalez, A. (2020). Dual language
learner supports for early learning settings (DLS). SRI Education.
Pakarinen, E., Lerkkanen, M. K., & von Suchodoletz, A. (2020). Teacher emotional support in relation to social
competence in preschool classrooms. International Journal of Research & Method in Education, 43(4), 444–460.
Perlman, M., Falenchuk, O., Fletcher, B., McMullen, E., Beyene, J., & Shah, P. S. (2016). A systematic review and
meta-analysis of a measure of staff/child interaction quality (the Classroom Assessment Scoring System) in
early childhood education and care settings and child outcomes. PLoS ONE 11(12), Article e0167660. https://doi.
org/10.1371/journal.pone.0167660
Pianta, R. C., DeCoster, J., Cabell, S., Burchinal, M., Hamre, B. K., Downer, J., LoCasale-Crouch, J., Williford, A., &
Howes, C. (2014). Dose–response relations between preschool teachers’ exposure to components of professional
development and increases in quality of their interactions with children. Early Childhood Research Quarterly,
29(4), 499–508.
Pianta, R. C., & Hamre, B. K. (2009). Conceptualization, measurement, and improvement of classroom processes:
Standardized observation can leverage capacity. Educational Researcher, 38(2), 109–119.
Pianta, R. C., Hamre, B. K., & Mintz, S. L. (2012). The CLASS-secondary manual. Teachstone.
Pianta, R. C., Hamre, B. K., & Nguyen, T. (2020a). Measuring and improving quality in early care and education. Early
Childhood Research Quarterly, 51, 285–287.
Pianta, R. C., La Paro, K. M., Payne, C., Cox, M. J., & Bradley, R. (2002). The relation of kindergarten classroom
environment to teacher, family, and school characteristics and child outcomes. The Elementary School Journal,
102(3), 225–238.
Pianta, R. C., Mashburn, A., Downer, J., Hamre, B., & Justice, L. (2008). Effects of web-mediated professional
development resources on teacher-child interactions in pre-kindergarten classrooms. Early Childhood Research
Quarterly, 23(4), 431–451.
Pianta, R. C., Whittaker, J. E. Vitiello, V., Ruzek, E., Ansari, A., Hofkens, T., & DeCoster, J. (2020b). Children’s school
readiness skills across the pre-K year: Associations with teacher-student interactions, teacher practices, and
exposure to academic content. Journal of Applied Developmental Psychology, 66, Article 101084.
Ramakrishnan, A., Zylich, B., Ottmar, E., LoCasale-Crouch, J., and Whitehill, J. (2021). Toward automated classroom
observation: Multimodal machine learning to estimate CLASS Positive Climate and Negative Climate. IEEE
Transactions on Affective Computing. https://doi.org/10.48550/arXiv.2005.09525
Reyes, M. R., Brackett, M. A., Rivers, S. E., White, M., & Salovey, P. (2012). Classroom emotional climate, student
engagement, and academic achievement. Journal of Educational Psychology, 104(3), 700–712.
Ritchie, S., Howes, C., Kraft-Sayre, M., & Weiser, B. (2001). Emerging Academic Snapshot. [Unpublished instrument].
University of California, Los Angeles.
Rimm-Kaufman, S. E., Curby, T. W., Grimm, K. J., Nathanson, L., & Brock, L. L. (2009). The contribution of children’s self-
regulation and classroom quality to children’s adaptive behaviors in the kindergarten classroom. Developmental
Psychology, 45(4), 958–972.
Rodgers, M. K., Hagler, N., Qiu, Y., Mathien, T., Park, J., & Faiello, M. (2020). Early learning performance funding
project: Cumulative evaluation final report, 2014–2019. Florida’s Office of Early Learning.
Rodgers, M. K., Qiu, Y., Leite, W., Hagler, N., Mathien, T., Schroeder, S., Reyes, C., Thorman, A., Glaser, L., & Fish, G.
(2018). Early learning performance funding project: Final evaluation report 2017–2018. Florida’s Office of Early
Learning.
Sabol, T. J., Soliday-Hong, S., Pianta, R. C., & Burchinal, M. R. (2013). Can rating pre-K programs predict children’s
learning? Science, 341(6148), 845–846. https://doi.org/10.1126/science.1233517
Sandilos, L., Goble, P., & Schwartz, S. (2020) Burnout and teacher–child interactions: The moderating influence of SEL
interventions in Head Start classrooms. Early Education and Development, 31(7), 1169–1185.
https://doi.org/10.1080/10409289.2020.1788331
Sandilos, L. E., Shervey, S. W., DiPerna, J. C., Lei, P, & Cheng, W. (2017). Structural validity of CLASS K-3 in primary
grades: Testing alternative models. School Psychology Quarterly, 32(2), 226–239.
Shin, H., & Ryan, A. M. (2017). Friend influence on early adolescent disruptive behavior in the classroom: Teacher
emotional support matters. Developmental Psychology, 53(1), 114–125.
Silinskas, G., Pakarinen, E., Lerkkanen, M-K., Poikkeus, A-M., & Nurmi, J-E. (2017). Classroom interaction and literacy
activities in kindergarten: Longitudinal links to Grade 1 readers at risk and not at risk of reading difficulties.
Contemporary Educational Psychology, 51, 321–335.
Slot, P. L., Boom, J., Verhagen, J., & Leseman, P. P. M. (2017). Measurement properties of the CLASS toddler in
ECEC in The Netherlands. Journal of Applied Developmental Psychology, 48, 79–91. https://doi.org/10.1016/j.
appdev.2016.11.008.
Smith, M. W. & Dickinson, D. K. (2002). Early Language & Literacy Classroom Observation (ELLCO) Toolkit. Brookes.
Soliday Hong, S. L., Sabol, T. J., Burchinal, M. R., Tarullo, L., Zaslow, M., & Peisner-Feinberg, E. S. (2019). ECE quality
indicators and child outcomes: Analyses of six large child care studies. Early Childhood Research Quarterly, 49,
202–217. https://doi.org/10.1016/j.ecresq.2019.06.009
Soukakou, E., (2016). Inclusive Classroom Profile, Research Edition. Brookes.
Sroufe, L. A. (1996). Emotional development: The organization of emotional life in the early years. Cambridge
University Press.
Stein, L., Therriault, S. B., Kistner, A., Auchstetter, A. M., & Melchior, K. (2016). Evaluation of level 4 school turnaround
efforts in Massachusetts: Part 1: Implementation Study. AIR.
Teachstone. (n.d.). Ventura county case study. https://info.teachstone.com/resources/case-studies/ventura
Vernon-Feagans, L., Mokrova, I. L., Carr, R. C., Garrett-Peters, P. T., Burchinal, M. R., & Family Life Project Key
Investigators. (2019). Cumulative years of classroom quality from kindergarten to third grade: Prediction to
children’s third grade literacy skills. Early Childhood Research Quarterly, 47, 531–540.
Vitiello, V. E., Bassok, D., Hamre, B. K., Player, D. W., & Williford, A. P. (2018). Measuring the quality of teacher child
interactions at scale: Comparing research-based and state observation approaches. Early Childhood Research
Quarterly, 44(33) 161–169.
Weiland, C., Ulvestad, K., Sachs, J., & Yoshikawa, H. (2013). Associations between classroom quality and children’s
vocabulary and executive function skills in an urban public prekindergarten program. Early Childhood Research
Quarterly, 28(2), 199–209. https://doi.org/10.1016/j.ecresq.2012.12.002
Wendling, B. J., Mather, N., LaForte, E. M., McGrew, K. S., & Schrank, F. A. (2015). Woodcock-Johnson IV test of early
cognitive and academic development. Boston. Houghton Mifflin Harcourt.
White, L. J, Fernandez, V. A., & Greenfield, D. B. (2020). Assessing classroom quality for Latino dual language learners
in Head Start: DLL-specific and general teacher-child interaction perspectives. Early Education and Development,
31(4). 599–627.
Williford, A. P., Maier, M. F., Downer, J. T., Pianta, R. C., & Howes, C. (2013). Understanding how children’s engagement
and teachers’ interactions combine to predict school readiness. Journal of Applied Developmental Psychology,
34(6), 299–309.
Wright, A., Farmer, D., & Kara, Y. (2018). Effects of sustained quality in pre-K, kindergarten, and first grade in Dallas
ISD. Center on Research and Evaluation, Southern Methodist University.
Workman, S. (2021). Promoting equitable access to quality child care. Center for American Progress. https://www.
americanprogress.org/article/promoting-equitable-access-to-quality-child-care/

PK3 2nded ReferenceManual v1.1 Virtual

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

PK3 2nded ReferenceManual v1.1 Virtual

Uploaded by

Copyright:

Available Formats

PRE-K – 3RD

Copyright © 2022 by Teachstone, Inc. All rights reserved.

Educators and Policymakers

Contributors, Consultants, and CLASS Advisory Board

The Path Ahead

CLASS 2nd Edition

What Has Changed

Center on Equity Increase Access Measure for Impact

Figure P.1. CLASS 2nd Edition Enhancements

Key Improvement Details

What is the CLASS CQI system?

CLASS Measurement Suite

What does CLASS measure?

Educators and children express little relational negativity verbally or nonverbally.

Educators are aware of and responsive to children’s needs—social, emotional, physical,

Educators support children’s growing behavioral regulation skills by creating

Content Time Number Engagement Format

What CLASS Does Not Measure

Theoretical and Empirical Foundations of CLASS

CLASS Activity Settings

How was CLASS developed?

Educator-child interactions can be reliably described,

CLASS and Child Outcomes

CLASS across a Diversity of Populations

Improving the Quality of Interactions

Targeted professional development helps educators

CLASS and the Office of Head Start

CLASS and QRIS

Systematic Improvements in Quality across Sectors

CLASS indicates where additional resources can be most

Ensuring More Equitable Distribution of Resources

Providing Data to Support CQI

Investing in High-Quality Improvement Programs

CLASS in Pre-K–12 Systems

CLASS and Educator Evaluation

CLASS and School Improvement

Establishing a CLASS Observation

CLASS Observation Components

Supporting Readiness in the Organization

Standard CLASS Observation Protocol

Adapting CLASS Observation Protocols to Align with Purpose

When to Who to How long

Live Observation Procedure

Video Observation Procedures

Coding Using Video

Conducting Informal CLASS Observations

Conducting CLASS Environment Observations

Maintaining a Pool of CLASS Observers

Recruiting CLASS Observers

Observers must be able to acknowledge their biases, learn strategies

Training CLASS Observers

Supporting CLASS Observers

Data Quality Checks

Providing Feedback to Educators

Reporting CLASS Data by Audience

Table 4.2. Calculating Classroom-Level Summary Scores

Step How Example

2. Calculate the reverse-coded Subtract the NC average from 8. NC average = 2

3. Calculate each domain Total the dimension averages for PC average = 6

Repeat for the remaining two 6 + 6 + 5.5 + 4 = 21.5

Promoting Equitable CLASS

Equitable Observation Practices

Recognizing and Reducing Bias in CLASS Observations