Use of Classroom Assessment Scoring System (CLASS™) in Head Start

The Classroom Assessment Scoring System (CLASS™) is an observation instrument that assesses the quality of teacher-child interactions in center-based preschool classrooms. CLASS™ includes three domains or categories of teacher-child interactions that support children's learning and development: Emotional Support, Classroom Organization, and Instructional Support. Within each domain are dimensions which capture more specific details about teachers' interactions with children.

Why is it important to assess the quality of teacher-child interactions?

The CLASS™ dimensions are based on developmental theory and research suggesting that interactions between children and adults are the primary way of supporting children's development and learning, and that effective, engaging interactions and environments form the foundation for all learning in early childhood classrooms.

How is CLASS™ scored and what do those scores mean?

CLASS is scored by trained and certified observers using a specific protocol. Following their observations of teacher-child interactions, CLASS™ observers rate each dimension on a 7-point scale, from low to high.

Scores of 1-2 mean the quality of teacher-child interactions is low. Classrooms in which there is poor management of behavior, teaching that is purely rote, or that lack interaction between teachers and children would receive low scores.

Scores of 3-5, the mid-range, are given when classrooms show a mix of effective interactions with periods when interactions are not effective or are absent.

Scores of 6-7 mean that effective teacher-child interactions are consistently observed throughout the observation period.

During the CLASS™ observation reviewers independently review and score each classroom using a computer-based scoring system in the Office of Head Start Monitoring Software (OHSMS). After the review OHSMS averages the scores across the grantee to result in grantee-level dimension scores. The dimension scores are then used to calculate the grantee-level domain scores. Reviewers do not have access to the grantee level score during or after the review.

What does the domain of Emotional Support include? What do classrooms that score in the high-range look like?

Emotional Support assesses the degree to which teachers establish and promote a positive climate in their classroom through their everyday interactions.

Classrooms that score well in this domain have teachers that are responsive to children, acknowledge children's feelings or emotions, help children resolve problems, redirect challenging behavior, and support positive peer relationships. Observations provide evidence that teachers and children support and respect one another. Teachers are aware of and respond to children's academic and emotional needs and consistently provide comfort, reassurance and encouragement. There is an emphasis on children's interests, motivations and points of view.

What does the domain of Classroom Organization include? What do classrooms that score in the high-range look like?

Classroom Organization assesses classroom routines and procedures related to the organization and management of children's behavior, time and attention in the classroom. High-scoring classrooms feature consistent schedules, well-designed learning centers, established routines, and sensitive and appropriate guidance strategies. Staff work together as a team. Classrooms with these characteristics give children a sense of stability and predictability that supports exploring, thinking about, and learning new things.

What does the domain of Instructional Support include? What do classrooms that score in the high-range look like?

Instructional Support assesses the ways in which teachers implement the curriculum to effectively promote cognitive and language development. This domain measures how teachers support and extend children's thinking, problem solving and conversational skills, and vocabulary. Effective teachers support children's engagement by making concepts and skills relevant to their everyday lives, asking questions that encourage children to analyze and reason, providing the right amount of help and offering feedback that acknowledges children's attempts.

Are there things that CLASS does not measure?

Yes. While effective interactions are critical and form the foundation for children's school success, they are only one piece of an effective early childhood program.  CLASS™ does not measure other important components of high quality teaching and learning such as the curriculum used, the process of the ongoing assessment of child progress, or individualized teaching.

The National Center on Quality Teaching and Learning (NCQTL) develops and disseminates staff development tools promoting evidence-based practices for improving preschool classroom teaching practices. These tools are designed to promote effective, engaging interactions and environments that research indicates are foundational for early learning.  The NCQTL resources align with CLASS™ dimensions, and help Head Start programs support classrooms that are well-organized and managed, provide social and emotional support, and demonstrate the instructional interactions and use of materials that stimulate children's thinking and skills.

Supporting local programs in their use of these tools is a cadre of Early Childhood Education (ECE) Specialists who are certified as CLASS™ trainers and who work directly on-site with local programs. The ECE specialists are available to local programs to present CLASS™ overviews or to train program staff to become CLASS™ observers. These specialists also conduct joint observations with Education Managers and Mentor Coaches for the purpose of assessing the professional development needs of teaching staff related to teacher-child interactions and then tailor training and technical assistance to the specific needs of that program.

Additionally, some local programs use their own training dollars to supplement the training and technical assistance received from NCQTL and the ECE specialists.

Section 641A(c)(2)(F) of the Head Start Act  (the Act) requires that the OHS monitoring review process include the use of a ''a valid and reliable research based observational instrument, implemented by qualified individuals with demonstrated reliability, that assesses classroom quality, including assessing multiple dimensions of teacher-child interactions that are linked to positive child development and later achievement.'' The Act also states, in Section 641(c)(1)(D), that such an instrument should be used as part of the system for designation renewal.

ACF consulted with leading early childhood assessment experts prior to selecting the instrument to be used. The experts agreed that CLASS™ was the instrument that best met the statutory requirement.  Ultimately, ACF selected CLASS™ Pre-K because it is an instrument that has been validated by over 10 years of research in educational settings. If you have additional questions, please direct them to DRS@headstartinfo.org.

Why do CLASS reviewers conduct two observations per classroom instead of the four recommended in the CLASS Manual?

OHS CLASS™ reviews are conducted for the purpose of obtaining a grantee-level rather than classroom-level score. To obtain a valid grantee-level score the University of Virginia (UVA) advised OHS to obtain a greater number of observation cycles across classrooms, rather than four cycles at the individual classroom level. Given this, ACF worked with the CLASS developers to determine the most appropriate number of observation cycles to be conducted. It was determined that two cycles across classrooms would be the most accurate reflection of a grantee-level score.

As a result, and in consultation with the CLASS™ developers, the number of observation cycles is two per each classroom observed. This permits the reviewer to conduct observations in more classrooms and so get a grantee-level score.

How are CLASS reviewers trained and their performance monitored?

CLASS-reliable reviewers are trained and certified as CLASS™ observers through Teachstone, the organization that provides training in CLASS. Teachstone, in collaboration with the Center for Advanced Study of Teaching and Learning (CASTL), a research center at the University of Virginia, grants certification of all CLASS™ observers or reviewers. Reviewers are required to attend a two-day Observation Training provided by a certified CLASS™ trainer and then pass a reliability test. This demonstrates that they are reliable with the CLASS™ Pre-K tool and are able to observe teacher-child interactions through the CLASS™ lens. Reviewers are required to recertify on an annual basis.

In addition, each review season CLASS™ reviewers are provided with professional development that includes activities such as refresher training conducted by a Teachstone trainer, webinars with CLASS™ experts, mentor coaching, and access to phone and email support. The quality assurance process includes periodic checks of reliability through a procedure known as calibration, which requires that reviewers code videos alongside another certified observer to ensure that the reviewer remains reliable.

How are the classrooms to be observed selected?

ACF has worked with statisticians to develop a statistically sound methodology for sampling the classrooms in which CLASS™ observations will be conducted. The sample of classrooms to be observed is computer generated and randomly drawn based on classroom data entered by the grantee into the Head Start Enterprise System (HSES). This sampling methodology results in a sufficient number of classes being selected from across the grantee to ensure that scores are representative of the grantee.

Once on site the CLASS™ reviewer is instructed to make every effort to maintain the original sample of classrooms. In a case where that is not possible, the CLASS™ reviewer receives guidance on how to choose a replacement.

Are there guidelines for when CLASS™ observations should not be conducted?

Yes. CLASS™ reviews are not to be conducted in the first two or last two weeks of the program year or during the winter holidays at the end of December. Grantees' classrooms, when the program is beginning and concluding its year or at the end of December, may not be representative of the classroom environment during the program year. Additionally, observations are not conducted during naptime or outdoor unstructured free play.

How are CLASS™ observations conducted in classes where children speak a language(s) other than English?

OHS requires that the CLASS™ reviewer be fluent in the predominant language used in the classroom. For example, Spanish language competency is evaluated for CLASS™ reviewers who will be assigned to conduct observations in programs and/or classrooms where Spanish is the dominant language spoken by the children. If there is not a reviewer available who is fluent in the language spoken in a classroom, a CLASS™ observation is not conducted.

What CLASS scores cause a grantee to be required to compete?

There are two circumstances under which a grantee is required to compete as the result of low CLASS™ scores. First, grantees with average CLASS™ scores below the established minimum on any of the three CLASS™ domains will be required to compete. These thresholds have been established as a score of 4 for the domain of Emotional Support, 3 for the domain of Classroom Organization, and 2 for the domain of Instructional Support.  Second, each year the ten percent of grantees reviewed that receive the lowest average scores in each domain are required to compete.

If a program scores in the bottom 10 percent of all Head Start programs, this means that the vast majority of Head Start programs were assessed at higher levels. However, if the lowest 10 percent in any of the three CLASS™ domains should include grantees with a score of 6 or 7, those grantees would not be required to compete, even if this means that fewer than 10 percent would be required to compete based on that domain.

What was the threshold for the lowest 10 percent of CLASS scores in 2012 by domain?

Grantees that had a review conducted in 2012 and that had scores less than or equal to the numbers below were in the lowest ten percent in each respective CLASS™ domain:

• Emotional Support – 5.4926
• Classroom Organization – 4.8571
• Instructional Support – 2.1923

In FY 2012, the average grantee-level scores received by Head Start programs during CLASS™ reviews were 5.90 in Emotional Support, 5.45 in Classroom Organization, and 2.98 in Instructional Support. Previous large-scale studies of CLASS™ have shown that the average preschool classroom scores are higher in the domains of Emotional Support and Classroom Organization than in the domain of Instructional Support.