Dear all,
I'm new to doing model comparison based on exceedance probability. I'm not sure I've been doing it right:
Say I include the models A, B, C, D. I find A is strongly the better model.
Then I include the models A, B, C, D, E. Now C is strongly the better model.
Is this normal? Should inclusion of E change whether A vs. C is better?
I would be very grateful for any pointers in the right direction..
Many thanks
Jacquie
|