Login  Register

suggestions for a suitable modelling approach, please!

classic Classic list List threaded Threaded
1 message Options Options
Embed post
Permalink
Reply | Threaded
Open this post in threaded view
| More
Print post
Permalink

suggestions for a suitable modelling approach, please!

P Mason
3 posts
Dear All,
This is probably simple, but I'm having difficulty conceptualising how to
proceed with a group of similar analyses. The data involve a dependent
variable that is derived from a number of others concerning whether an
individual cites a particular aspect of their neighbourhood (eg vandalism,
crime, dogs, traffic etc) as being a problem. The dependent variable is the
total number of aspects cited as being a problem (out of a maximum of 10). I
want to investigate how this total varies with respect to several
independent variables. As you would expect, the distribution of the totals
is highly skewed - on average people might claim 2 problems, say, but
smaller numbers of people will claim up to all 10. I keep envisaging this as
a problem of "frequencies within frequency classes" but am not getting much
further in my attempts to decide what type of GLM I should be using. Any
ideas gratefully received.
Thanks,
Phil Mason