| Special education teachers certified through the University of Wisconsin-Madison Department of Rehabilitation Psychology and Special Education were surveyed on the use of charting and graphing of student performance data. Item generation for a survey instrument was conducted (1) in collaboration with special education teachers, and (2) using Davis and Salasins' (1975) AVICTORY theoretical model of program implementation as a guide for the development of contextual items addressing classroom ecology.;Of 99 respondents to the survey, 84.9% indicated the use of some type of charting, while 55.6% indicated the use of graphing. Thirty-five respondents (35.4%) indicated the use of frequent data collection (at least twice a week per curriculum area, in most curricular areas of teaching). Results depicted teachers' overall positive attitude toward frequent data collection while indicating concern with issues such as effect on accountability and lack of time needed for implementation.;Reliability and cluster analyses of 29 contextual items indicated that (a) items grouped to reflect AVICTORY factors did not perform as internally consistent factors, and (b) a factor-like structure did not emerge from the data. Data from each contextual item were subsequently entered stepwise in a discriminant analysis. Six items predicted the reported use or nonuse of frequent data collection with 75% accuracy. These items represented "built-in materials", "peer teachers as a source of help" (a negatively weighted predictor), "extent to which peer teachers would value data collection", "usefulness for multidisciplinary team and IEP uses" (a negatively weighted predictor), "a big change in practice", and data collection as a "need for effective teaching".;In view of negative coefficients obtained in the discriminant analysis, caution in the interpretation of the results was stressed. The results may have been due to the effect of predictor variables on the use or nonuse of frequent data collection, the effect of use or nonuse on the predictor variables, or the result of significant influences not addressed in the survey. |
