metricTester (version 1.2.2)

sesOverall: Overall per simulation-null-metric SES test

Description

This function provides one of many ways of summarizing and considering simulation results.

Usage

sesOverall(simulation.list, test, concat.by)

Arguments

simulation.list

A summarized results list such as one output from reduceResults(). See examples.

test

Either "ttest" or "wilcotest", depending on whether the user wants to run a two-sided t-test or a Wilcoxon signed rank test.

concat.by

Whether randomizations were concatenated by richness, plot or both.

Value

A data frame summarizing the mean, overall standardized effect sizes and the significance of those devations from expectations for each simulation, null, metric combination. This test works across all iterations, and looks for overall shifts in SES from expectations (see details for for expectations).

Details

This function provides one way of summarizing and considering simulation results. It takes as input a vector of all standardized effect sizes for all plots from a given simulation-null-metric combination, and calculates the mean of the vector and whether it differs significantly from a mean of zero. It does this either with a simple two-sided t-test, or with a Wilcoxon signed rank test. If the latter, and if there are three different spatial simulations with names random, filtering and competition, the test is two-sided, less and greater, respectively.

References

Miller, E. T., D. R. Farine, and C. H. Trisos. 2015. Phylogenetic community structure metrics and null models: a review with new methods and software. bioRxiv 025726.

Examples

#not run
#results <- readIn()
#summ <- reduceResults(results)
#examp <- sesOverall(summ$ses, "both")