rsmodel is a basic fitting function for rating scale models.
rsmodel(y, weights = NULL, start = NULL, reltol = 1e-10,
deriv = c("sum", "diff"), hessian = TRUE,
maxit = 100L, full = TRUE, ...)rsmodel returns an S3 object of class "rsmodel",
i.e., a list with the following components:
a named vector of estimated item-specific parameters (without the first item parameter which is constrained to 0) and estimated cumulative relative treshold parameters (again without first threshold parameter which is also constrained to 0),
covariance matrix of the parameters in the model,
modified data, used for model-fitting, i.e., cleaned for items without
variance, centralized so that the first category is zero for all items
and without observations with zero weight. Be careful, this is different than for
objects of class "raschmodel" or "btmodel", where
data contains the original data,
logical vector of length ncol(y), which
indicates which items have variance (TRUE), i.e., are identified and have been
used for the estimation or not (FALSE),
integer vector of length ncol(y), which
contains the number of categories minus one per item,
number of observations (with non-zero weights),
original number of observations in y,
the weights used (if any),
logical indicating whether the data contains NAs,
list of elementary symmetric functions and their derivatives for estimated parameters,
log-likelihood of the fitted model,
number of estimated parameters,
convergence code from optim,
number of iterations used by optim,
tolerance passed to optim,
original function call.
item response object that can be coerced (via as.matrix)
to a numeric matrix with scores 0, 1, ... Typically, either
already a matrix, data frame, or dedicated object of class
itemresp.
an optional vector of weights (interpreted as case weights).
character. If "sum" (the default), the first derivatives of the elementary symmetric functions are calculated with the sum algorithm. Otherwise ("diff") the difference algorithm (faster but numerically unstable) is used.
an optional vector of starting values.
logical. Should the Hessian of the final model be computed?
If set to FALSE, the vcov method can only return NAs
and consequently no standard errors or tests are available in the
summary.
further arguments passed to optim.
logical. Should a full model object be returned? If set to FALSE,
no variance-covariance matrix and no matrix of estimating functions are computed.
rsmodel provides a basic fitting function for rating scales models,
intended as a building block for fitting rating scale trees. It
estimates the rating scale model in the parametrization suggested by
Andrich (1978), i.e., item-specific parameters \(\xi_{j}\) who mark
the location of the first absolute threshold of an item on the theta axis and
cumulative relative threshold parameters \(\kappa_{k}\) are
estimated by the function rsmodel.
rsmodel returns an object of class "rsmodel" (and
class "pcmodel") for which several basic methods are available,
including print, plot, summary, coef,
vcov, logLik, discrpar, estfun,
itempar, threshpar, and personpar.
Andrich D (1978). Application of a Psychometric Rating Model to Ordered Categories Which Are Scored with Successive Integers. Psychometrika, 2(4), 581--594.
pcmodel, gpcmodel, raschmodel,
nplmodel, btmodel
o <- options(digits = 4)
## Verbal aggression data
data("VerbalAggression", package = "psychotools")
## Rating scale model for the other-to-blame situations
rsm <- rsmodel(VerbalAggression$resp[, 1:12])
summary(rsm)
## visualizations
plot(rsm, type = "profile")
plot(rsm, type = "regions")
plot(rsm, type = "curves")
plot(rsm, type = "information")
plot(rsm, type = "piplot")
options(digits = o$digits)
Run the code above in your browser using DataLab