| Abstract: | General conditions which should be met in the development of the idea of science processes and the potential benefits which would result are suggested. An approach to the definition of science processes based on variables and variable handling is outlined. It is argued that this should occur specifically in the context of educational practice. The processes should be elucidated within a taxonomy of tasks, the ultimate components of which should be interpretable and transferable skills. The development and utilization of such a taxonomy can most obviously occur in the context of assessment. The techniques of item banking and domain definition in domain‐referenced tests are then discussed in relation to the above issues. Multi‐dimensional labelling of an item bank is considered to be appropriate, and suggestive of a research programme for investigating the requirements of transferability and interpretability. The labelling scheme resulting from an attempt to identify the significant dimensions of the age 15 Assessment of Performance Unit (APU) science item bank is reported. The paper indicates the effectiveness of the scheme and discusses the value of labelling those items within the assessment programme which draw heavily on the conceptual framework of taught science. A discussion of how the labelling scheme/item bank can be used in an attempt to identify the skills required in undertaking tasks based on science processes, and thus to develop tests of such identifiable skills for diagnostic and placement purposes in schools, concludes this paper. |
