The main use of this function is for identification of outliers.
Usage
## S3 method for class 'resid.grouped'
summary(object, K = 2, observed = NULL, ...)
Arguments
object
an object of class resid.grouped.
K
the cutoff point to identify outliers
observed
a numeric vector of possible observed data, e.g., the mean of
the interval in which the true data lie.
...
additional arguments; currently none is used.
Details
Taking into advantage the realizations of the standardized residuals r_i provided by the Multiple Imputation
scheme, we can estimate the probability
Pr(|r_i| > K), i = 1,…,n,
which can be regarded as the
probability of the ith sample unit being an outlier.
Value
a numeric matrix with columns, the fitted values, the estimated residuals, and the percentage of each sample unit
having an absolute residual greater than the value given by K. If !is.null(observed) its value is given as
the first column of the matrix.