Computes effective standard deviation intervals for observed assay results. These intervals have at least 68.27 percent coverage.
# S3 method for assaytest
predict(object, newdata, ...)
An object of class "assaytest"
A vector of observed values. If missing, uses object$x.
additional arguments
A data frame with the observed values, lower, and upper confidence limits
Takes the Umax
element from the assaytest
object and treats it as the known precision
parameter. For the constant SD model, the effective standard deviation interval for observed data
value y is (y-Umax, y+Umax). For the constant CV models the effective SD interval uses either
normConstCVCI
(for the "normal" model) or lognormConstCVCI
(for the "lognormal" model).
Although Umax
is an upper bound (not an estimate) of the precision parameter, simulations have shown
that treating Umax
as the true precision parameter
gives effective SD intervals with coverage of at least 68.27 percent (see Fay, Sachs, and Miura, 2016).
Fay, MP, Sachs, MC, and Miura, K (2016). A Hypothesis Testing Framework for Validating and Assay for Precision (unpublished manuscript).