TEACHERS EVALUATION OF CORRELATIONAL REASONING SKILLS

Two experiments sought to increase our understanding about teachers' p roficiency in assessing higher order thinking skills. In Experiment 1 eight grade 7 teachers participated in training and subsequently asses sed samples of student performance from grades 5, 7, 9, 11, and OAC (O ntario Academic Credit) on two forms of a correlational reasoning test . Teachers' ratings of the 100 samples of student performance did not correspond highly with an expert's assessment although the novice asse ssors did show sensitivity to variation in student performance due to student age and test difficulty. In Experiment 2, 31 grade 7 and 8 tea chers participated in an in service program on correlational reasoning that required them to implement an instructional unit in their own cl assrooms. They were subsequently trained to assess student performance and marked a sample of pre- and posttest data from their own students . The prediction that more intense implementation would enhance teache rs' understanding of correlational reasoning concepts and concomitantl y, their assessment skills was not supported. Teachers' assessment pro ficiency was higher than in Experiment 1, but did not correspond highl y with the standard set by the expert marker. Sensitivity to student g rowth was evident, however. Results are discussed in terms of teachers ' purposes for doing assessment and implications for research.