July 12, 2011 — It's a sunny Wednesday morning at Malcolm W. Cole Child Care Center in Charlottesville. Preschoolers are settling back into their classrooms after a fire drill. Angie Mundy leads toddlers in a one-on-one finger-painting project, talking with each about the colors they are smearing across construction-paper canvases.

Across the room, Tanja Doder leads other youngsters in several rounds of "Old MacDonald Had a Farm." As she holds up a cat figure, a distinct meow erupts from the group. "Kitty cat, yes!" Doder responds.

In this brightly lit room, cheerfully painted and covered with colorful posters, the center's staff groups together nine children ages 16 to 24 months in one room in compliance with Virginia codes and maintains the proper teacher-to-toddler ratios. Mundy and Doder have years of experience in early child care settings and strictly follow safety procedures.

Adherance to environmental factors in a child care setting are important; however, the features that matter most to children's healthy academic and social development are high-quality, one-on-one interactions, according to early childhood research – those like Mundy and Doder engage in as they model language, encourage children to use language and challenge children to think critically.

To ensure all preschool students receive such care and instruction, states are increasingly examining classroom relationships. The classroom observation instrument many consider to be an indispensible tool for assessing these interactions was developed at the University of Virginia's Center for Advanced Study of Teaching and Learning, and is called the Classroom Assessment Scoring System, or CLASS.

"It's the single most effective thing I've ever been involved with for children and staff," said Mary Margaret Gardiner, child care quality program manager at Children, Youth and Family Services, which oversees the Malcolm Cole Center and 23 others in the Charlottesville area. The centers all participate in the Virginia Star Quality Initiative, created in 2006 under Gov. Timothy M. Kaine's Working Group on Early Childhood Initiatives.

Gardiner and her staff have been trained to use the CLASS instrument to rate each of the centers, and use the information gathered to develop a two-year, quality-improvement plan and provide mentoring with the aim of raising future scores.

"People always have the best intentions with children, but not always the best practice," she said. "They love children, but they need support. With the CLASS tool, you can see what's going on. You get a good picture of a child's day-to-day experience."

Gardiner finds that centers most commonly need support in the area of instructional interactions, such as modeling language and encouraging children to learn new words. "It takes a different way of thinking about talking to children," she explained. She pointed to research findings that 85 percent of language in child care settings is directive – "Do this. Don't do that."

Her goal is to help child care center directors understand that "the relationship between the provider and the child is what drives honest-to-God quality. You have to see what the child is thinking, draw them out and ask questions," she said. "You must be one-on-one, sitting beside the child or with the child on your lap and following the child's lead." 

According to 2005 data from the National Household Education Surveys Program, 57 percent of children ages 3 to 5, and 20 percent of infants and toddlers, spend time in center-based care. The vast majority of these child care centers had little oversight, especially of their efforts to improve children's school readiness, until 13 years ago, when a few states began developing quality rating systems. The Star Quality Initiative is Virginia's version.

"We wanted to organize and coordinate what is a fairly fragmented system of early learning," said Kathy Glazer, who directed the Working Group on Early Childhood Initiatives. "We needed a mechanism to better organize early learning programs and find ways to increase quality."

The Star Quality Initiative, now in its fourth year and administered jointly by the Office of Early Childhood Development and the Virginia Early Childhood Foundation, addresses four areas, or standards: education, qualifications and training of staff; staff-to-child ratios and group size; program environment and instructional practices; and interactions between children and their teachers and peers.

Child care centers that volunteer to participate in the initiative are assigned between one and five stars, based on observed and documented outcomes in each area.

As the Virginia working group members were designing the rating system, they took into account the latest research about how to have the most impact on children's learning and best prepare them for school, Glazer said.

"We found high-quality and effective interactions to be a critical piece," she said, "so we wanted to home in on this particular feature of classroom quality." In fact, they chose to weight the interactions score more heavily, so that it accounts for one-third of a program's total rating.

CLASS was developed and tested in more than 6,000 classrooms by U.Va. Curry School of Education researchers. It was not initially designed as a tool for monitoring the quality of child care centers, said senior scientist Bridget Hamre, although it has been adapted and assessed for this purpose over the last five years.

Hamre is a key member of the research team led by Curry School Dean Robert C. Pianta, who directs CASTL. About two decades ago the team began identifying a set of teacher-student interactions – those related to emotional and instructional support as well as to classroom management – that seemed related to students' future levels of achievement. They then developed a coding system, which became the CLASS instrument, to guide individuals in characterizing what they observe in classrooms for research purposes.

Initially focused on pre-kindergarten through third-grade classes, the researchers have been able to characterize the quality of teaching in classrooms across the country and have found evidence to connect the quality of interactions to student outcomes.

"For example, we know from the research that when first-grade children at risk for school failure are exposed to classrooms with effective teacher-student interactions, they perform on par with peers not at risk, both socially and academically," Hamre said.

In each domain of interest, the CLASS instrument describes the full range of possible teacher-child interactions from what is considered to be highly effective to ineffective. "When we identify and measure interactions," Hamre said, "we can then create opportunities to promote more effective interactions through teacher education and professional development."

Currently, 24 states have implemented quality rating and improvement systems, and an additional 22 are in some phase of development. In addition to Virginia, the CLASS tool is being used in rating systems in Arizona, Minnesota, Montana and Massachusetts, according to Lourdes Lambert of Teachstone Inc., the company that administers training for CLASS observers. In Iowa, Kansas, Maine, Rhode Island, Washington and Wyoming, CLASS is included in the mix of allowable and recommended measures that local jurisdictions can choose, and it is being considered by eight other states as they plan for rating-system development.

As interest grows, states are also seeking to assess teacher-child interactions at younger levels, which prompted the U.Va. researchers to develop a toddler version of CLASS – one for use in classrooms like that of Doder and Mundy, Hamre said, that addresses a related set of appropriate interactions for children age 15 to 36 months. More recently, an infant version addressing interactions with children from birth to 18 months has been developed and is still being piloted.