About this srcd poster session

---

Panel information
Panel 18. School Readiness/Childcare

Abstract

---

The Classroom Assessment Scoring System (CLASS) measures the elements of children’s learning experiences that research demonstrates matter most—daily interactions among educators and children. Multiple studies have documented the ways in which interactions as measured by CLASS support children’s development (Perlman et al., 2016; Wang et al., 2020). Moreover, CLASS is used widely for quality measurement and improvement across the United States such as the Office of Head Start, Quality Rating and Improvement Systems, and K-12 school systems. However, much of this research has occurred at the preschool level. One reason for the lack of studies in infant and toddler settings may be due to the current structure of the Infant CLASS (Hamre et al., 2014) and Toddler CLASS (La Paro et al., 2012). These tools are separate instruments with distinct structures in terms of number of domains and dimensions measured. This separation often creates difficulty in the field, including misalignment of pre and post observation data, confusion about operationalization of interactions across age groups, and user burden for deciding which tool to use in mixed-age settings. Additionally, the current tools insufficiently address the diversity of effective interactions seen in infant and toddler care. This presentation will discuss the development of a combined CLASS Infant-Toddler tool, including revisions to better capture interactions in a variety of settings, and the initial tests of its use in the field. Two studies were conducted to assess the initial validity and reliability of the CLASS 2nd Edition Infant - Toddler: a comparison of videos scored on the current and new tools, and an implementation pilot to gather observer feedback on the use of the tool. In the comparison study, 8 observers coded 289 video cycles of 51 infant, toddler, or mixed-age classrooms. In the implementation pilot, 13 observers observed 54 classrooms (Head Start, private, and family child care settings) for a total of 216 cycles. Together these studies indicate that the new tool demonstrates sufficient face validity, construct validity, internal consistency, and concurrent validity. Specifically, external expert reviews and observer feedback suggests the tool can capture effective interactions across the age groups in a developmentally appropriate way and can be used to support educator interactions in a continuous quality improvement model. A confirmatory factor analysis demonstrated that a 2-factor model best fit the data (chi-square = 36.73, p < .0001, RMSEA = 0.07), with a good fit and all dimensions loading significantly onto their corresponding factors. The internal consistency of the combined tool was similar to the current tools (Chronbach’s ɑ = .89), with an increased alpha for the two factors from the current Toddler CLASS. There were no significant differences in means across the three tools - Infant, Toddler, and 2nd Edition Infant-Toddler. Finally, correlations among the tools indicate high levels of agreement. Early validation research suggests that additional validation with more information on educator and child outcomes will help to provide further psychometric evidence. This research will have key implications for early care and education policy such as quality thresholds and equitable measurement of interactions.

Author information

Author	Role
Victoria L. Kintner-Duffy, Ph.D., Teachstone	Presenting author
Sarah Caverly, Ph.D., Teachstone	Non-presenting author
Jamie DeCoster, Ph.D., University of Virginia	Non-presenting author

⇦ Back to session