fence.by.group: Fence Values Based on Group Means and Standard Deviations

View source: R/fence.by.group.R

fence.by.groupR Documentation

Fence Values Based on Group Means and Standard Deviations

Description

Identifies values of x that are a certain number of standard deviations from the mean within a particular set of grouping variables and replaces them with that boundary value (e.g., values more than 3 standard deviations from the mean are replaced with the mean +/= 3 standard deviations). For example, fence response times that are more than 2.5 standard deviations from the mean within each cell of a PrimeType x WordFrequency factorial experimental design.

Usage

fence.by.group(x, INDEX, sds = 3, print = TRUE, na.rm = FALSE)

Arguments

x

a numeric vector.

INDEX

list of one or more grouping variables, typically factors, each of the same length as x.

sds

number of standard deviations away from the mean at which values should be replaced.

print

a logical value indicating whether to display the number and percentage of replaced elements.

na.rm

a logical value indicating whether NA values should be ignored when computing the mean and standard deviation.

Value

vector in which values more than the specified number of standard deviations from each cell mean have been replaced with the boundary value.

See Also

flag.by.group to identify such observations without replacing them.

trim.by.group to treat such observations as missing data rather than replacing them with a new value.

fence to fence values a certain number of standard deviations away from the overall mean rather than a group mean.

Examples

data(sleepstudy, package='lme4')
sleepstudy$Reaction.Fenced <- fence.by.group(sleepstudy$Reaction,
  sleepstudy$Subject, sds=3) # fence RTs 3 std devs from subject mean

data(VerbAgg, package='lme4')
VerbAgg$Anger.Fenced <- fence.by.group(VerbAgg$Anger,
  list(VerbAgg$btype, VerbAgg$situ), sds=3)
# fence 3 std devs from behavior type x situation type cell mean

sfraundorf/psycholing documentation built on April 23, 2022, 2:50 a.m.