We are new users of SPSS Modeler and are looking at 3 years' worth of survey
data in a CHAID model to predict withdrawal from college. Our dataset is
6300 - is this too small to run a partition? Our accuracy increases by 1%
with a partition but the predictive factors change depending on whether or
not we use a partition.
--
Sent from:
http://spssx-discussion.1045642.n5.nabble.com/=====================
To manage your subscription to SPSSX-L, send a message to
[hidden email] (not to SPSSX-L), with no body text except the
command. To leave the list, send the command
SIGNOFF SPSSX-L
For a list of commands to manage subscriptions, send the command
INFO REFCARD
Locked
