Using Q-mode factor analysis in biostratigraphic ordination

Although traditionally restricted to the analysis of closed compositional data, the Imbrie Q-mode method appears also to be of utility when applied to binary data sets common in biostratigraphy. Its advantages include its utilization of the tally concept, and its ability to analyze very large numbers of species simultaneously. When the structure of the data set is simple, the method appears to be capable of ranking samples along a temporal gradient, or coenocline, defined by the second unroated factor. Varimax and oblique rotations do not appear to increase the efficacy of this ranking. Even when facies effects are important, a temporal ranking of samples can be achieved, providing that the variation due to time is greater than that due to facies.