|
|
Dear List,
I am trying to do some spline regressions on some data from the Survey of
English Housing. I've had a look at the SPSSX-L archives and can understand
how
you would do this using REGRESSION (if the knots are fixed) or NLR/CNLR if
you
also want to estimate the location of the knots (which I probably do).
However,
these are based on a continuous response variable, whereas, in my
case, the dependent variable is dichotomous (eg is vandalism a problem in
your
neighbourhood? Yes/No) so I need to adapt the general method to suit a
logistic
model. Does anyone have any ideas as to how to do this in SPSS? Or am I
barking up entirely the wrong tree?
Thanks,
Dr Phil Mason
Department of Urban Studies
University of Glasgow
|
Locked
