calcPW: Compute pair-wise comparisons estimates and bootstrap...

Description Usage Arguments Value

View source: R/calcPW.R

Description

Calculates the mean or median sample estimates of all pair-wise differences within a factor. Can be set compute the maximum Bayes factors (MBFs) and indicate which of the comparisons are statistically significant after controlling the false discovery rate, if the number of bootstrap repetition (B) is set to greater than 0.

Usage

1
calcPW(data, calcStat, alpha, B, fdr, method = "bca")

Arguments

data

dataframe. Contains numeric variable named 'resp' for responses and character variable named 'group' containing factor levels. c

calcStat

function olr character. Bootstrap statistic function. Functions other than mean or median used at own risk.

alpha

numeric. 1-alpha is the coverage of the confidence intervals.

B

integer. Number of bootstrap samples per subgroup. If equal to zero, then the MBF is not calculated.

fdr

numeric. False discovery rate.

method

'bca' for BCa or 'perc' for percentile. Bootstrap confidence interval and p-value method.

Value

A tibble of the comparison groups and estimates of differences in calcStat, as well as the confidence intervals and MBFs if bootstrapping was performed.


MiguelRodo/ggboot documentation built on May 1, 2018, 12:22 a.m.