The methodology (i.e. procedures) used in testwork for modifying, improving, selecting, designing or developing processes is reviewed and possible improvements described. Minerals and particles could be characterised more comprehensively with modifications of existing procedures._x000D_
Identification of the dominant phenomenon in a process could help bring the results of laboratory tests closer to plant performance. Improvements in pilot plant testwork could reduce the frequency and magnitude of problems in materials handling and separation in production processes, and increase the quantity and quality of the data obtained from a limited number of tests. Longer-term improvements in methodology depend on some changes in basic research._x000D_
Information is required about all of the phenomena that occur in a production process rather than sophisticated models of specific phenomena. Theories, correlations and computer simulations could be used more effectively in testwork if they were simple, capable of forecasting the effects of changes and applicable to a number of similar processes.