\name{varDescription}
\alias{varDescription}
\alias{toString.varDescription}
\alias{print.varDescription}
\title{
Variable description object.
}
\usage{
varDescription(name=character(0),nodeName=character(0),
          role=character(0),levels=character(0), isOrdered = TRUE)
toString.varDescription(x, ...)
print.varDescription(x, ...)
}
\description{
  A meta-data description of a StatShop variable.  Usually created by
  reading an Assessment Model Description file.

  Note:  StatShop currently only supports discrete variables.
}
\arguments{
  \item{name}{External identifier for the variable (used by Evidence
    Identification Process).}
  \item{nodeName}{Internal name of the variable (defaults to same as
    name).}
  \item{role}{A string code identifying the role of the variable.  One
    of \sQuote{SM}---Student model
    variable,---\sQuote{Observable}---observable outcome variable,---or
    \sQuote{EM}---internal Evidence/Link model variable (e.g., context
    effect).}
  \item{levels}{A string vector giving the names of the possible states
    of the variable.  Note:  if states are ordered assumes that the
    highest state is first.}
  \item{isOrdered}{A logical value indicating whether or not the
    elements should be regarded as ordered.}
  \item{x}{A \code{varDescription} we are printing.}
  \item{\dots}{Arguments passed to print function.}
}
\value{
  A class with elements as described in the arguments section.
}
\details{
  StatShop assumes states are ordered from highest to lowest, e.g.,
  \code{High},\code{Medium}, \code{Low}.

  StatShop also allows variables to have a different internal name
  (\code{nodeName}) used in the graphs and external name (\code{name})
  used in data requests send to the scoring engine (e.g., pretest data
  files).  
}
\references{
  \url{http://research.ets.org/~ralmond/StatShop/dataFormats.html}
}
\seealso{
  \code{\link{parseVar}},\code{\link{xmlDiscreteVariable}}
}
\examples{
  Theta <- varDescription("theta","theta",role="SM",
          levels=c("Advanced","Proficient", "Basic", "Below Basic"))
  X2 <- varDescription("X2","ScoredResponse",role="Observable",
                      levels=c("Correct","Incorrect"))
  Context <- varDescription("Context","Context",role="EM",
                      levels=c("Familiar","Unfamiliar"))
}
\keyword{interface}