Function to estimate the Grubbs test statistic.
Usage
grubbs.test(x, ...)
# S3 method for default
grubbs.test(
x,
var.index = 1,
replicate.index = 2,
material.index = 3,
laboratory.index = 4,
data.name = NULL,
alpha = 0.05,
...
)
# S3 method for lab.qcdata
grubbs.test(x, alpha = 0.05, ...)Arguments
- x
an R object (used to select the method). See details.
- ...
arguments passed to or from methods.
- var.index
Scalar with the column number corresponding to the observed variable (the critical to quality variable). Alternatively, a string with the name of a quality variable can be provided.
- replicate.index
Scalar with the column number corresponding to the index each replicate.
- material.index
Scalar corresponding to the replicated number.
- laboratory.index
Scalar that defines the index number of each laboratory.
- data.name
String specifying the name of the variable which appears on the plots. If name is not provided, it is retrieved from the object.
- alpha
The significance level (0.05 for default)
References
Wilrich Peter-T. (2013), Critical values of Mandel's h and k, the Grubbs and the Cochran test statistic. Asta-Advances in Statistical Analysis, 97(1):1-10.
ASTM E 691 (1999), Standard practice for conducting an interlaboratory study to determine the precision of a test method. American Society for Testing and Materials. West Conshohocken, PA, USA.
Examples
library(ILS)
data(Glucose)
Glucose.qcdata<- lab.qcdata(Glucose)
str(Glucose.qcdata)
#> Classes ‘lab.qcdata’ and 'data.frame': 120 obs. of 4 variables:
#> $ x : num 41 41.5 41.4 41.2 42 ...
#> $ replicate : Factor w/ 3 levels "1","2","3": 1 2 3 1 2 3 1 2 3 1 ...
#> $ material : Factor w/ 5 levels "A","B","C","D",..: 1 1 1 1 1 1 1 1 1 1 ...
#> $ laboratory: Factor w/ 8 levels "Lab1","Lab2",..: 1 1 1 2 2 2 3 3 3 4 ...
#> - attr(*, "data.name")= chr "Glucose"
grubbs.test(Glucose.qcdata)
#>
#> Test Grubbs
#>
#> Critical value: 1.154478
#>
#> Alpha test: 0.00625
#> Material Gmax G.max p.value.max Gmin G.min p.value.min
#> 1 A Lab8 1.909229 0.0489 Lab7 1.915242 0.0485
#> 2 B Lab4 1.564273 0.0809 Lab1 1.490219 0.0899
#> 3 C Lab4 2.984235 0.0102 Lab7 1.387137 0.1040
#> 4 D Lab8 2.388179 0.0241 Lab7 2.423819 0.0229
#> 5 E Lab2 1.576391 0.0795 Lab7 1.551749 0.0823