Odd looking residual plot - not sure what transform to use if any

Asked Aug 20 '15 at 06:52

Active Aug 20 '15 at 07:28

Viewed 121 times

I am concerned about the residual plot shown. The (count) data are over-dispersed, with about 40% 0s, median is 2, maximum is 300 or so.

I am not sure what how to proceed with this - it is not exactly like any of the classic examples that are on various teaching websites, and I have not yet found a transformation that makes the plot look "well-behaved".

edited Aug 20 '15 at 07:28

Dawny33

2,239
1
21
37

asked Aug 20 '15 at 06:52

Vincent Laufer

1

what model are these residuals coming from? – Christoph Hanck Aug 20 '15 at 06:56
@ChristophHanck the command used first is: model – Vincent Laufer Aug 20 '15 at 06:57
3

have you tried a count data model if your data is count data? – Christoph Hanck Aug 20 '15 at 07:01
yes, library(MASS) model – Vincent Laufer Aug 20 '15 at 07:02
3

That almost straight line at the bottom must come from those almost 40% zeros, not much to do about that! – kjetil b halvorsen Aug 20 '15 at 07:46
1

Closely related: [Interpreting plot of residuals vs. fitted values from Poisson regression](http://stats.stackexchange.com/questions/25068). Another possibility (although likely not appropriate in this particular instance) is discussed at [How to model this odd-shaped distribution](http://stats.stackexchange.com/questions/49443). – whuber Aug 20 '15 at 14:28

Odd looking residual plot - not sure what transform to use if any

0 Answers0