Dear All,
I have a sample of discrete, non-negative count data where the response
variable is the number of misprints a typist makes on a page.
Many of the observations have a value of 0 so the data cannot be transformed
to a normal distribution. And as some typists make many misprints, the
variance was much larger than the mean, so a Poisson model could not be
fitted either.
I would like to find some way of testing to see if there is a difference
between certain typists and the number of misprints they make.
It has been suggested to fit a negative binomial model to the data. How
would I go about doing this, and could a regular testing method be set up to
test whether new typists are faring better than the current typists.
Any help, or references to information would be much appreciated.
Amy Chaffin
[log in to unmask]
_________________________________________________________________
Chat with friends online, try MSN Messenger: http://messenger.msn.com
|