-- Readme file : Supplement 4 - Anderson and Walsh (2013). This supplement contains .csv files of the results of the simulation study for the Ekofisk and Norway simulations. The definitions of the variables are given below: id : An identification number for the parameter set within the simulation. main_seed : The random number seed used to create the random number seeds for each parameter set within the simulation. seed : The random number seed used for random number generation for each parameter set. name : Name of the simulation. distribution : Statistical distribution of the simulated data (Bernoulli, log-normal, or negative-binomial). transformation : Transformation applied to simulated data before testing. Possible values include: "loge1" - natural logarithm of data plus 1; "quarter-power" - quarter power transformation; "chisq" - chi-squared transformation; "no" - no transformation. distance : The distance/dissimilarity measure used. Possible values include: "bray-curtis0" - Bray curtis dissimilarity; "euclidean" - euclidean distance; "jaccard" - Jaccard dissimilarity. dispersion : Flag to indicate whether the negative binomial dispersion is fixed across groups ("fixed"), or whether it varies as the mean of group 2 changes ("varies"). p : Number of variables (dimensions). g : Number of groups. nsim : Number of simulations performed for each parameter set. nperm.pam : Number of permutations performed for the PERMANOVA, ANOSIM, and Mantel tests. nperm.d : Number of permutations performed for the PERMDISP test. n_# : The sample size of group number #. N : The total number of samples across all groups. Delta : The proportion that the mean of group 2 is between the mean of the first location (group 1) and the second location. For example, if the simulation concerns location A and B of the Ekofisk data set, and Delta=0.5, then the mean of group 1 is the mean of location A, and the mean of group 2 is halfway (arithmetic average) between the mean of location A and location B. method : Which test was performed. Possible values include : ANOSIM, Mantel, PERMANOVA, PERMDISP.c, and PERMDISP.p. PERMDISP.c is the PERMDISP test calculated with the classical normal approximation. PERMDISP.p is the PERMDISP test calculated using permutation of residuals. t1 : Rate of rejection of H0 over the nsim simulations for the parameter set. The alpha level is 0.05, so this is the proportion of p-values less than or equal to 0.05 out of nsim simulated data sets. t1.lb : The lower bound of t1 based on the binomial distribution. t1.ub : The upper bound of t1 based on the binomial distribution.