Dear allstat,
I am analysing some data on tuberculosis declaration rate in a region of
Switzerland.
I am comparing different groups of the population and modelling the trend in
time by means of Poisson regression.
I have the total number of cases which have been declared yearly in the last
30 years in the population under study. Since there is no sampling I do not
know how to interpret the confidence intervals that I obtain. Should I
ignore them for the period of observation? Can I use them to make hypotheses
on future trends?
Could somebody help me further?
Thanks in advance,
With best regards,
Monica Daigl Cattaneo
--
Monica Daigl Cattaneo
Limmattalstrasse 1
CH-8049 Zürich
Switzerland
|