Determining the number of interacting species: significant factor analysis

The technique of significant factor analysis is described. This technique yields statistically valid criteria for defining the minimum number of linearly independent entities in a array of data. Analyses of examples from th literature are compared with other methods for extracting this information. Analysis of the subunit dissociation of hemoglobin illustrates the extent and the limits to model-independent conclusions that can be drawn. The technique is readily applicable with programs available in standard statistical packages.