Paleostatistics

From Wikipedia, the free encyclopedia

Paleontology often faces phenomena so vast and complex they can be described only through statistics.

First applied to the study of a population in 1662 statistics is today a basic tool for natural sciences practitioners, and a solid acquaintance with methods and applications is essential for communication purposes within the scientific community.

Thanks to the diffusion of powerful low-cost computers and the availability of many software tools for statistical analysis, data elaboration is now open to a much wider users pool than before.

Statistics offers to paleontology the tools needed to describe and summarize data (base statistics -- average, standard deviation, distributions), to stress and characterize relations existing between two sets of data, with reference to one or more taxonomic groups (correlation analysis, multiple regression, cluster analysis) and finally allows the testing of ipotheses and the development of new ipotheses from the available data (factor analysis, correspondence analysis).

A general skill in applying these few methods is enough to set up a basic analysis of both quantitative or semi-quantitative data, as a complement to a traditional palaeontological research. Statistical analysis alone on the other hand does not prove anything and its worth is directly dependent on the quality of the data used. Adopting a statistical approach to the data does not push back the paleontologist, and to the countrary turns the paleontologist's experience into the one essential component in a well-developed statistical analysis.