geepack (version 1.1-6)

ordgee: GEE for Clustered Ordinal Responses

Description

Produces an object of class `geese' which is a Generalized Estimating Equation fit of the clustered ordinal data.

Usage

ordgee(formula = formula(data), ooffset = NULL, id, waves = NULL,
       data = parent.frame, subset = NULL, na.action = na.omit,
       contrasts = NULL, weights = NULL, z = NULL,
       mean.link = "logit", corstr = "independence",
       control = geese.control(...), b = NA, alpha = NA,
       scale.fix = TRUE, scale.val = 1, int.const = TRUE,
       rev = FALSE,...)

Arguments

formula

a formula expression as for glm, of the form response ~ predictors. See the documentation of lm and formula for details. As for glm, this specifies the linear predictor for modelling the mean. A term of the form

ooffset

vector of offset for the odds ratio model.

id

a vector which identifies the clusters. The length of `id' should be the same as the number of observations. Data are assumed to be sorted so that observations on a cluster are contiguous rows for all entities in the formula.

waves

an integer vector which identifies components in clusters. The length of waves should be the same as the number of observation. components with the same waves value will have the same link functions.

data

an optional data frame in which to interpret the variables occurring in the formula, along with the id and n variables.

subset

expression saying which subset of the rows of the data should be used in the fit. This can be a logical vector (which is replicated to have length equal to the number of observations), or a numeric vector indicating which observation numbe

na.action

a function to filter missing data. For gee only na.omit should be used here.

contrasts

a list giving contrasts for some or all of the factors appearing in the model formula. The elements of the list should have the same name as the variable and should be either a contrast matrix (specifically, any full-rank matrix with as ma

weights

an optional vector of weights to be used in the fitting process. The length of weights should be the same as the number of observations.

z

a design matrix for the odds ratio model. The number of rows of z is $$c^2 \sum n_i(n_i - 1)/2,$$ where $n_i$ is the cluster size, and $c$ is the number of categories minus 1.

mean.link

a character string specifying the link function for the means. The following are allowed: "logit", "probit", and "cloglog".

corstr

a character string specifying the log odds. The following are allowed: "independence", "exchangeable", "unstructured", and "userdefined".

control

a list of iteration and algorithmic constants. See geese.control for their names and default values. These can also be set as arguments to geese itself.

b

an initial estimate for the mean parameters.

alpha

an initial estimate for the odds ratio parameters.

scale.fix

a logical variable indicating if scale is fixed; it is set at TRUE currently (it can not be FALSE yet!).

scale.val

this argument is ignored currently.

int.const

a logical variable; if true, the intercepts are constant, and if false, the intercepts are different for different components in the response.

rev

a logical variable. For example, for a three level ordered response Y = 2, the accumulated indicator is coded as (1, 0, 0) if true and (0, 1, 1) if false.

...

further arguments passed to or from other methods.

Value

• An object of class "geese" representing the fit.

References

Heagerty, P.J. and Zeger, S.L. (1996) Marginal regression models for clustered ordinal measurements. JASA, 91 1024--1036.

See Also

glm, lm, geese.

Examples

data(respdis)
resp.l <- reshape(respdis, varying =list(c("y1", "y2", "y3", "y4")),
                  v.names = "resp", direction = "long")
resp.l <- resp.l[order(resp.l$id, resp.l$time),]
fit <- ordgee(ordered(resp) ~ trt, id=id, data=resp.l, int.const=FALSE)
summary(fit)

data(ohio)
ohio$resp <- ordered(as.factor(ohio$resp))
fit <- ordgee(resp ~ age + smoke + age:smoke, id = id, data=ohio)
summary(fit)