Hi all,
I have come across an interesting question on proportional significance testing (Z tests).
I would like to compare a proportion of people who are positive to an index compared to another group. The index is a composite of 5 variables (questions)). The index is formed by calculating the proportion of positive responses for all the questions for each respondent.
e.g. if 30 respondents answered the 5 index questions this would be a total of 5*30= 150 answers. If 75 of these 150 answers were positive then the index would be 50%.
So my question is should I use n=30 or n =150 for the Z test calculation.
Your help on this would be greatly appreciated.
Thanks,
Ann
You may leave the list at any time by sending the command
SIGNOFF allstat
to [log in to unmask], leaving the subject line blank.
|