Brian
Add to your list - a factor that is not considered in the model that has a
large impact on the response. Like your two examples, this should also
appear as a gap (or gaps) in run charts or in 2 variable scatterplots.
In addition highly rounded data can create some strange looking residuals
plots.
Max Porter
"Brian Mcmillan" <[log in to unmask]>
Sent by: "A UK-based worldwide e-mail broadcast system mailing list"
<[log in to unmask]>
12-Sep-2006 22:52
Please respond to "Brian Mcmillan" <[log in to unmask]>
To
[log in to unmask]
cc
Subject
Gap in residuals plot
Hello all,
Does anyone know what might cause a gap in the middle of a residuals plot?
(predicted values plotted against residuals)
I'm guessing it could be one of two things - a range of outcome values
that never occur, or a range of predictor varibles that never occur,
possibly caused by combining two data sets measuring variables that have
very different ranges?
Hmmm.....
Any thoughts welcome! Thanks in advance,
Brian
|