Dear List Members
I have available published summary statistics (mean, s.d. and n) which have
been calculated from observations in a population-based study. There is a
separate set of statistics for each gender/ethnic group (six sets in all).
I do not have access to the original data.
Intuitively I guess I am able to calculate an "expected group mean" for any
specific gender/ethic mix, by simply manipulating the six means ((mean1 *
n1) + (mean2 * n2) (etc to 6)) / (sum n1 to n6)
Now the query - I would be grateful for advice (or reference) on how to
calculate an expected s.d. for the resulting "expected mean".
Thanks in anticipation
Neil S.
[log in to unmask]
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|