interlabstats: Statistics for interlaboratory studies

In qmrobust: Quality Management: Examples from the concrete industry

Description

Calculates statistics for interlaboratory studies. interlabstats is a generic function.

interlabstats.formula generates the statistics based on a simple analysis of variance.

interlabstats.default uses a suitable fitted model object, currently either an lmeMod or a rlmeMod object, and extracts the desired statistics from it.

Usage

interlabstats(object, ...)
## Default S3 method:
interlabstats(object)
## S3 method for class 'formula'
{formula, data, subset}
## S3 method for class 'interlab'
interlabstatsprint(x, groups=TRUE, ...)

Arguments

object	for interlabstats.default: object of class lmerMod or rlmerMod, obtained from function lmer of package lme4 or rlmer of package robustlmm, respectively.
formula	Formula of type y ~ group , where y is the measurement and group is the laboraotory factor
data	dataset containing the variables used in formula
subset	usual subset argument, passed to model.frame
x	object of class interlab to be printed
groups	for print.interlab: should statistics for groups be printed (if available)?
...	further arguments to be passed to the methods or to print.default

Details

By interlab.formula, the estimated quantities are obtained from a classical fixed effects Analysis of Variance.

Value

A list containig

mean	overall mean
sigma	within-group standard deviation
siggroup	between-group standard deviation
repeatability	standard error of a difference between measurements of the same lab
reproducibility	same, for different labs
groups	table, containing number of observations, mean and standard deviation for each group
data	the two relevant variables of the original data.frame
call	the call to the function
method	estimatrion method. "classical" for interlabstats.formula, class(object) otherwise.

Argumentx of print.interlab:

x	interlab object, as generated by interlabstats
groups	Should group table be printed?
...	Further arguments passed to format

Author(s)

Werner A. Stahel, ETH Zurich

References

Stahel ...

See Also

plot.interlab, lmer, rlmer

Examples

data(d.perm)
dd <- d.perm[d.perm$section=="w.1",]
dd[dd$team=="A","perm.log"] <- dd[dd$team=="A","perm.log"] - 0.3
  ## otherwise, there is no between group variation
( r.cl <- interlabstats(perm.log~team, data=dd) )
print(r.cl, groups=FALSE)

## Mixed models: lme
require(lme4)
r.lme <- lmer(perm.log~(1|team), data=dd, na.action=na.omit)
summary(r.lme)
interlabstats(r.lme)

## robust
require(robustlmm)
r.rlme <- rlmer(perm.log~(1|team), data=dd, na.action=na.omit)
summary(r.rlme)
interlabstats(r.rlme)

qmrobust documentation built on May 2, 2019, 6:50 p.m.