Testing For Trend In C-Section rates Using Proc Catmod

Reply
Occasional Contributor
Posts: 11

Testing For Trend In C-Section rates Using Proc Catmod

Hi,

I want to test for the above trend using Proc Catmod. I have 'age', 'indication', 'type of hospital' and 'source of payment' as my 'designs-effects' variable. I have both the frequencies (counts) and the rates for C-Sections in the State of New England for the 6 years years 2000 - 2005.

I need help regarding the following:

1. What variable should I use as the option for the 'weight' statement? The 'count' of C-Sections or their actual 'rate'?

2. Since I'm using the weighted least squares approach, I'll include the option wls in my model statement. What option should I use for the response statement?

3. I intend to treat the variable 'year' as the response effect. Is this the right approach? If not, do you have an alternative suggestion?

Any additional advice is most welcome.

Thanks.
Ask a Question
Discussion stats
  • 0 replies
  • 251 views
  • 0 likes
  • 1 in conversation