Proportions comparison

[kende jan]

Dear all,

I want to compare  two proportions of  disease in two populations : group 1 (1200/15000) and group 2 (260000/6500000). However  I would take into account the number of physicians involved in each group G1 (1600 physicians) and G2 (1.6 million). Please can someone can help me ?

 

Thanks

 

[Tom]

Besten Dank für Ihr Mail. Ich bin abwesend und werde erst ab dem 14. Februar meinen Posteingang wieder regelmässig anschauen und mich dann so schnell wie möglich melden. Mit freundlichen Grüssen Thomas Balmer

[lucameyer]

Are the group samples or populations? If they are sample you could run a Hypothesis Test for Difference Between Proportions, but if they are populations I do not think you need to run any test.

Concerning the number of physicians's base, again: if you have samples of larger group you *could* run hypothesis testing - depending on your substantive research questions, but if you have populations probably all you can do is to compute other indexes - i.e. (15000-1200)/1600 vs (6500000-260000)/16000000 to see if the G1's physicians are possibly more effective than G2's physicians to help with disease's prevention/cure.

HTH,
Luca