tionship of observational measures to child outcomes, especially in classroom-type early care and educational settings, and to a lesser extent in home-based care (Bryant, forthcoming; Burchinal, forthcoming; Burchinal et al., 2008). In addition, observational measures are used in evaluation studies to assess whether an intervention to improve practice in home-based or center-based early care and educational settings has affected caregiver/teacher practice or overall quality (for example, Bryant, 2007, and Pianta, 2007). An observational measure designed to assess parenting skills as a tool in caregiver or teacher professional development or for formative assessment should be detailed and descriptive so that it can help to direct improvement. In contrast, a measure used for research, summative assessment, or for accountability purposes, even if detailed, should be easily summarized in quantifiable ratings, so that scores can be compared over time and across settings.
Purposes, in turn, have implications for who conducts the observation. If the goal is professional development or formative assessment, observations might be done by individuals directly involved. For example, observations of parenting skills might be done by a home visitor; a child care program teacher or administrator could do observations of early care and education. If summative assessment or accountability is a goal, it is preferable that observation measures be administered by someone who is not directly connected to the program being evaluated, although program staff may sometimes perform this role if sufficient safeguards are in place to ensure the reliability and validity of the observations.
Most existing measures assess the social environment well and the learning environment at a very general level, but only a few adequately assess practices designed to teach academic or social skills specifically. Development of observational measures is just beginning to catch up with the increased political emphasis on academic preparation in programs for young children (National Institute for Early Education Research, 2006). We summarize below some existing observational measures of the home and center-based environments, without attempting to be exhaustive. For all of these measures, there is some evidence for