Advantages to and Challenges of Using Ratings of Observed Teacher-Child Interactions 468 views

This dissertation presents three independent studies that are linked in their focus on issues relevant to observing teacher-child interactions. The first study confirms the degree to which observed teacher-child interactions predict children's academic and social outcomes, and the second two studies inform the challenges in maintaining this predictive validity when using observations in large-scale contexts. Study 1 illustrates how children exposed to high quality teacher-child interactions consistent with effective classroom organization and instructional support in both pre-kindergarten and kindergarten scored significantly higher on assessments of their language and literacy skills. Study 2 examines the relationship between observed scores using the Classroom Assessment Scoring System (CLASS; Pianta, La Paro, and Hamre, 2008) and observation protocol characteristics, specifically the day of the week, month of the year, and duration of an observation cycle in minutes. CLASS scores were stable across these factors, with a few exceptions for the CLASS domain of Classroom Organization. Study 3 describes the extent of rater calibration resulting from a large-scale training effort by the Office of Head Start and explores rater characteristics that predict calibration. The majority of raters trained in this large-scale effort passed an initial calibration assessment and rater beliefs predicted the degree of calibration. Collectively, these three studies demonstrate that observational assessment can provide meaningful information about teachers and children in large-scale contexts despite challenges faced in planning and implementation. Note: Abstract extracted from PDF text

PHD (Doctor of Philosophy)

teacher-child interactions; ratings

English

All rights reserved (no additional license for public reuse)

2010-08-01