首页|Corrigendum to “Distances to compositional equilibrium” [J. Geochem. Explor. 227 (2021) 106793] (Journal of Geochemical Exploration (2021) 227, (S0375674221000716), (10.1016/j.gexplo.2021.106793))
Corrigendum to “Distances to compositional equilibrium” [J. Geochem. Explor. 227 (2021) 106793] (Journal of Geochemical Exploration (2021) 227, (S0375674221000716), (10.1016/j.gexplo.2021.106793))
扫码查看
点击上方二维码区域，可以放大扫码查看
原文链接
NSTL
Elsevier
? 2021 Elsevier B.V.The authors regret that in Section 5, “Discovering stoichiometric equilibrium” (Egozcue et al., 2021), some misleading statements and notations were detected. These mistakes are placed in a paragraph referred to testing the hypothesis of equilibrium of the olivine data set used in Section 4.2. A corrected paragraph is here reproduced for easy reading. A final comment is also added. Corrected paragraph There are circumstances in which distances to a possible equilibrium may not be small enough to accept that the sample is in equilibrium. A statistical test on H0: E(δeq) = 0, where E denotes expectation, can be carried out. In fact, δeq is a real random variable. Maintaining δeq for denoting the sample values, both the sample mean [Formula presented] and standard deviation S(δeq) can be obtained from the sample. Moreover, the univariate distribution of [Formula presented] can be approximately normal if the sample size is large enough. The standard t-student test, or alternatively a Chi-square test, on the mean value is then appropriate. If the distribution of [Formula presented] cannot be assumed normal, bootstrap techniques are also available. This kind of testing equilibrium has been largely studied for Hardy-Weinberg equilibrium, in particular, in a compositional approach see (Graffelman and Egozcue, 2011). For instance, the olivine sample used in Section 4.2 can be checked for olivine equilibrium in Equation (8) using the t-student statistic for testing [Formula presented] where n = 8827 is the sample size. The test statistic is t = 4.4027 and the p-value is 1.082e?05. This suggests rejection of H0, indicating that the stoichiometric equilibrium of olivine in the sample can be rejected. The reasons for the rejection of equilibrium of the olivine data set are found in the fact that it is the result of collecting several data sets producing an heterogeneity in the samples. Also the test is referred to Mg-Fe olivine, excluding other possible substitutions beyond the Mg-Fe pair, which may produce outliers. The authors would like to apologise for any inconvenience caused.
NSTL
Elsevier
