Statistical inferential testing

Information on the constituents of rocks, sediments and soil necessarily derive from samples; it is never complete. Inferences drawn from data are uncertain, and the main aim of statistical testing is to attach probabilities to those inferences. The tests are prefaced by hypotheses or models and are designed to refute the hypotheses or determine the probabilities that the data do not comply with the models. A result is declared statistically significant when the probability, P, is less than some small value, conventionally P < 0.05. Statistical significance is not the same as scientific significance, and calculated values of P should always be presented along with standard errors in reports and papers so that clients and readers can draw their own inferences from results.