Statistical analyses of scatterplots to identify important factors in large-scale simulations, 1: Review and comparison of techniques

Procedures for identifying patterns in scatterplots generated in Monte Carl o sensitivity analyses are described and illustrated. These procedures atte mpt to detect increasingly complex patterns in scatterplots and involve the identification of (i) linear relationships with correlation coefficients, (ii) monotonic relationships with rank correlation coefficients, (iii) tren ds in central tendency as defined by means, medians and the Kruskal-Wallis statistic, (iv) trends in variability as defined by variances and interquar tile ranges, and (v) deviations from randomness as defined by the chi-squar e statistic. A sequence of example analyses with a large model for two-phas e fluid flow illustrates how the individual procedures can differ in the va riables that they identify as having effects on particular model outcomes. The example analyses indicate that the use of a sequence of procedures is a good analysis strategy and provides some assurance that an important effec t is not overlooked. (C) 1999 Elsevier Science Ltd. All rights reserved.