ERROR-WEIGHTED MAXIMUM-LIKELIHOOD (EWML) - A NEW STATISTICALLY BASED METHOD TO CLUSTER QUANTITATIVE MICROPALEONTOLOGICAL DATA

The advent of readily available computer-based clustering packages has created some controversy in the micropaleontological community concer ning the use and interpretation of computer-based biofacies discrimina tion. This is because dramatically different results can be obtained d epending on methodology. The analysis of various clustering techniques reveals that, in most instances, no statistical hypothesis is contain ed in the clustering model and no basis exists for accepting one biofa cies partitioning over another. Furthermore, most techniques do not co nsider standard error in species abundances and generate results that are not statistically relevant. When many rare species are present, st atistically insignificant differences in rare species can accumulate a nd overshadow the significant differences in the major species, leadin g to biofacies containing members having little in common. A statistic ally based ''error-weighted maximum likelihood'' (EWML) clustering met hod is described that determines biofacies by assuming that samples fr om a common biofacies are normally distributed. Species variability is weighted to be inversely proportional to measurement uncertainty. The method has been applied to samples collected from the Fraser River De lta marsh and shows that five distinct biofacies can be resolved in th e data. Similar results were obtained from readily available packages when the data set was preprocessed to reduce the number of degrees of freedom. Based on the sample results from the new algorithm, and on te sts using a representative micropaleontological data set, a more conve ntional iterative processing method is recommended. This method, altho ugh not statistical in nature, produces similar results to EWML (not c ommercially available yet) with readily available analysis packages. F inally, some of the more common clustering techniques are discussed an d strategies for their proper utilization are recommended.