plot.robust.design(x, y = NULL, fun = mean, data.pts=T, ...)
REQUIRED ARGUMENTS:
x
an object of class
"robust.design".
OPTIONAL ARGUMENTS:
y
the response(s). If a response is not given, all non-factor
components of
x will be the response(s).
fun
the function to be applied to the data in each factor level.
data.pts
logical flag for whether data points are plotted.
print.values
a logical flag for whether mean values will be added to plot.
...
arguments passed to plot.
SIDE EFFECTS:
A version of a main effects plot is produced on the current graphics device.
DETAILS:
Experiments using robust design methods are usually intended to detect
factors that influence the variance and/or the mean. The vertical
lines for each factor level indicate the smallest and largest
observations; the lines joining the levels show the change in the
mean. The plot is designed to draw attention to the range of
observations within each level of a factor (as a surrogate for the
variance) in addition to the change in mean.