biasedPlots	Demonstrates the biased plots that result from F-statistic
  based feature selection with data where there are known groupings,
  followed by calculation and plotting of discriminant function scores.
  [The 'accuracies' are also biased; use \code{defectiveCVdisc()} and/or
  \code{demo(classifyRandom())} to demonstrate the effect]
CVscoreplot	Demonstrates the methodology used by \code{cvscores()} to
  obtain scores (plotted using \code{scoreplot()}) that, because derived
  from the scores for the successive 'test' sets at the successive steps
  of the cross-validation, minimize bias.
classifyRandom  Shows, for random normal data, how the biases in the
  training set accuracy measure and in the cross-validation accuracy
  measure when all of the data has been used to select the 'best' features
  change with number variables selected.  Accuracy estimates from the use of
  cross-validation, with re-selection of features at each cross-validation
  fold, are shown for comparison.
