Description
Analysis of mass and volume changes associated with prograde regional metamorphism is a complex, tedious, and time intensive process. In an attempt to make such a task less time intensive G. H. Girty wrote a Matlab 5.0, menu-driven program named Statistical Analysis of Compositional Data (SACD) which removes much of the tedium from such analyses. However, this program is a i version, and prior to this study had not been tested. Any new area of research focused on the statistics of elemental mass change and/or the power of the Student's t test in resolving mass change would be appropriate areas to test the usefulness of SACD. One of the most modern investigations of mass change in pelitic mudstones brought on by Barrovian metamorphism is the recently completed study of the Wepawaug Schist of Connecticut by Ague ( 1994 ), who utilized the Student's t test to evaluate whether or not elemental mass in the pelitic mudstones of the Wepawaug Schist were changed during prograde regional metamorphism. Though the study by Ague (1994) has received much acclaim, the tatistical power of the Student's t test to recognize elemental mass differences as small as those calculated by Ague (1994) had not been attempted prior to this study. The results of this study indicate that SACD produces reliable data in a user-friendly format in a considerably shorter time period than does such an analysis using Excel and Mathematica. In addition, data indicate that in going from the biotite-chlorite zone to the garnet zone that Ca and Sr were removed from the Wepawaug Schist resulting in a change in rock mass of -10 :t: 4 grams/ 100 grams. In going from the garnet to the staurolite zone, Si and Na were lost while K was added, resulting in a total change in rock mass of -8 ± 5 grams/ 100 grams. In going from the staurolite to the kyanite zone, Ba appears to have been lost but there is no statistically detectable change in the total rock mass. These conclusions are supported not only by the results of the Student's t test but also by power values that are greater than 80%. However, power for all other elements analyzed by Ague ( 1994) falls below 80%, and , as a result, there is little statistical data that warrant acceptance of the null hypothesis that no change in mass has occurred for these nuclides. Thus, a consideration of power leads to some refinement of the general conclusions reached in the benchmark work by Ague (1994).
