PREDICTION OF CLASSROOMS THAT ARE AT RISK: IMPLICATIONS FOR STAFF DEVELOPMENT (ARIZONA).

Abstract

This study was designed to isolate variables that could predict classrooms where there is some risk of low academic achievement. Observations were conducted in 18 second through sixth grade classrooms in Tucson, Arizona. A minimum of three observations were conducted in each classroom for each of two subject areas, reading and mathematics. The California Achievement Test was used as the pre and post outcome measure. Multiple correlation analyses were used to predict classrooms that would be expected to have low academic achievement gains over the course of one year of instruction. These "at-risk" classrooms were identified using variables derived from a model of academic learning time. The results indicated that the best predictors of reading achievement were process variables such as focusing on task and receiving corrective feedback. This was true when both achievement gain and achievement residuals were time variables such as allocated and engaged time. Though further research is necessary, this study suggests that variables associated with a model of academic learning time appear promising as predictors of classrooms that are at-risk. The possibility of early and valid predictions of this kind has obvious implications for staff development programs. Using the data from classroom observations, specific inservice procedures could be used to alleviate the causes of academic risk. Staff development programs targeted to classrooms that are identified as at-risk would also allow more efficient use of scarce inservice dollars.