getContrasts {gnm}R Documentation

Estimated Contrasts and Standard Errors for Parameters in a gnm Model

Description

For each set in a specified list of sets of parameters from a gnm model, computes the estimated simple contrasts (i.e., differences) with the first parameter in the set, and estimated standard errors for those estimated differences. Where possible, quasi standard errors are also computed.

Usage

getContrasts(model, sets = NULL, nSets = 1, dispersion = NULL,
  use.eliminate = TRUE, ...)

Arguments

model a model object of class "gnm".
sets a vector of indices or a list of such vectors. If NULL, a Tk dialog will open for parameter selection.
nSets the number of vectors of indices to use when sets is NULL.
dispersion either NULL, or a positive number by which the model's variance-covariance matrix should be scaled.
use.eliminate logical; see vcov.gnm
... arguments to pass to other functions.

Details

The indices must all be in 1:length(coef(object)). If sets = NULL, a Tk dialog is presented for the selection of indices (model coefficients).

For each set of coefficients selected, differences with the first coefficient and their standard errors are computed. A check is performed first on the estimability of all such differences.

If sets is non-NULL, the value of the nsets argument is ignored.

Value

EITHER a list (normally of length nSets) of objects of class qv — see qvcalc — (when sets is a list, or when sets is NULL and nSets > 1);
OR an object of class qv (otherwise).

Author(s)

David Firth

References

Firth, D (2003). Overcoming the reference category problem in the presentation of statistical models. Sociological Methodology 33, 1–18.

Firth, D and Menezes, R X de (2004). Quasi-variances. Biometrika 91, 65–80.

See Also

gnm, se, checkEstimable, qvcalc, ofInterest

Examples

set.seed(1)
data(yaish)

## Fit the "UNIDIFF" mobility model across education levels -- see ?yaish
unidiff <- gnm(Freq ~ educ*orig + educ*dest +
               Mult(Exp(educ), orig:dest),
               ofInterest = "[.]educ", family = poisson,
               data = yaish,  subset = (dest != 7))
## Examine the education multipliers (differences on the log scale):
unidiffContrasts <- getContrasts(unidiff, ofInterest(unidiff))
plot(unidiffContrasts,
  main = "Unidiff multipliers (log scale): intervals based on
           quasi standard errors",
  xlab = "Education level", levelNames = 1:5)

[Package gnm version 0.9-2 Index]