Factorial Design and Statistical Analysis of Flotation Experiments

'Trial and error' and 'one-variable-at-atime' type of conventional experiments are inadequate to study a complex process like flotation, with results dependent on a number of interacting variables. Recent studies (1 – 3) suggest that factorial experiments, evolved by Sir Ronald Fisher in late 1920's, offer several advantages in flotation investigations involving quantitative factors.However, fac-torial design did not become popular in flotation studies, and so far it has not been tried as an alternative to the 'trial and error' method. Therefore, the study of graphite fines flotation is presented in this paper to illustrate the advantages of factorial design over the conventional 'trial and error' approach.