asymdhd: Difference asymmetry function for two dependent groups or a...

In GRousselet/rogme: Robust Graphical Methods For Group Comparisons

Difference asymmetry function for two dependent groups or a single distribution

Description

asymdhd computes a difference asymmetry function for one or more distributions. The difference asymmetry function provides perspective on the degree a distribution is symmetric about zero, by quantifying the sum of q and 1-q quantiles. If the groups do not differ, then the difference scores should be symmetric about zero. In particular, the sum of q and 1-q quantiles should be zero. If the distribution is symmetric the function should be approximately a horizontal line. If in addition the median of the difference scores is zero, the horizontal line will intersect the y-axis at zero. Confidence intervals and p values are returned for each quantile sum. The FWE is controlled via Hochberg's method, which is used to determine critical p values based on the argument alpha. To plot the results use plot_diff_asym.

Usage

asymdhd(
  data = df,
  formula = obs ~ gr,
  qseq = seq(5, 40, 5)/100,
  alpha = 0.05,
  nboot = 1000,
  todo = NULL,
  doall = FALSE
)

Arguments

data	A data frame in long format. One column is a factor describing the groups; another column contains the values/observations for each group. A properly formatted data frame can be created using mkt1 or mkt2. Missing values are not allowed.
formula	A formula with format response variable ∼ predictor variable, where ~ (tilde) means "is modeled as a function of".
alpha	Expected long-run type I error rate - default = 0.05.
nboot	Number of bootstrap samples - default = 1000.
todo	A list of comparisons to perform - default = NULL.
doall	Set to TRUE to compute all comparisons - default = FALSE. Not executed if a todo list is provided.
q	Quantiles to estimate - default = 0.05:0.05:.4 - must be <.5.

Value

A list of data frames, one data frame per comparison. Each data frame has one row per decile. The columns are:

• Column 1 = quantiles

• Column 2 = quantiles of differences

• Column 3 = 1 - quantiles of differences

• Column 4 = sum of quantiles

• Column 5 = lower bounds of the confidence intervals

• Column 6 = upper bounds of the confidence intervals

• Column 7 = critical p_values based on Hochberg's method

• Column 8 = p_values

Note

This function combines Rand Wilcox's difQpci and Dqdif R functions, from Rallfun-v32.txt - see https://github.com/nicebread/WRS/, http://dornsife.usc.edu/labs/rwilcox/software/.

References

Wilcox, R.R. & Erceg-Hurn, D.M. (2012) Comparing two dependent groups via quantiles. J Appl Stat, 39, 2655-2664.

See Also

hd asymhd for independent groups

Examples

set.seed(21) # generate data
n <- 100 # sample size per condition
C1 <- rnorm(n) # condition 1
C2 <- C1 + rnorm(n) + 2 # condition 2
# Data with 3 independent groups and 2 dependent conditions per group
library(tibble)
df <- tibble(gr = factor(c(rep("group1",n),rep("group2",n),rep("group3",n))),
  cond1 = c(C1, C1+rnorm(n), C1+rnorm(n)),
  cond2 = c(C2, C2 + 1, C2 + 3) ) # make tibble
library(dplyr)
df <- mutate(df, diff = cond1 - cond2)

out <- asymdhd(df, diff ~ gr) # use the default parameters
out <- asymdhd(df, diff ~ gr, alpha = .90) # specify alpha
out <- asymdhd(df, diff ~ gr, nboot = 500) # specify the number of bootstrap samples
out <- asymdhd(df, diff ~ gr, todo = list("group1", "group2")) # specify list of comparisons
out <- asymdhd(df, diff ~ gr, q = seq(.1,.4,.1)) # specify the quantiles
out <- asymdhd(df, diff ~ gr, doall = TRUE) # compute all tests

GRousselet/rogme documentation built on Nov. 12, 2022, 4:38 a.m.