4 4 re 3 Pr Science and Mathematics Education Centre The Effectiveness of Predict-Observe-Explain Technique in Diagnosing Students’ Understanding of Science and Identifying Their Level of Achievement Chong-Wah Liew This thesis is presented for the degree of Doctor of Science Education of Curtin University of Technology October 2004DECLARATION ‘This thesis contains no material which has been accepted for the award of any other degree or diploma in any university. To the best of my knowledge and belief this thesis contains no material previously published by any other person except where due acknowledgement has been made. Signature vo hy pae: 15" Uet 200ABSTRACT The purpose of this research program was to explore the effectiveness of the Predict- Observe-Explain (POE) teaching/leaming technique to diagnose students’ understanding of science and identify students’ level of achievement with reference to the Science Student Outcome Statements for Australian schools. ‘This research employed an interpretive action research approach with a sample of students from three Australian metropolitan high schools in grades 9, 10, 11, and 12, whose ages ranged between 14 and 17 years, Three data collection methods were used to generate data for interpretation, namely, written POE responses of students, in-class journals and student interviews. Data collected were interpreted using three theoretical perspectives, namely, Chi et al.’s theory of ontological categories, Hewson and Hennessey’s conceptual change theory to determine the epistemological status of students’ understanding of science, and Chinn and Brewer's model to classify types of students’ responses to contradictory observations. This purpose of using this methodology was to obtain an in-depth, plausible and credible account of students’ understanding and their level of achievement. POE tasks were concerned with heat and the expansion of water, solubility of salt, and power and resistance of light globes. The data revealed common ideas amongst. students that are contrary to scientists’ science; furthermore, students showed that they were able to articulate their own ideas based on the POE tasks. ‘The findings in this research reveal that these POEs were effective in capturing a range of possible student observations and prediction outcomes when worded in an open-ended format. Quality information on students’ understanding and on the way they responded to contradictory data was obtained when POEs were administered by teacher demonstrations and were designed to produce phenomena that were clear, immediate and had only one aspect to observe. Furthermore, the data suggest that POEs are effective in identifying students’ achievement across levels within a substrand of the Australian Student Outcome Statements and enable the teacher to observe and document a spread of achievement over a range of levels rather that a