A couple of questions that I've been stuggling with:
1) Testing proportions with small samples and finite populations
Can anyone tell me how to test/produce CI's for the difference between two proportions calculated from two small (around 20-30)
independent samples taken from two finite populations (about 100). The samples are rather small for the normal approximation to
be accurate plus there is the finite population correction to include (or does this make it better?).
2) Confidence levels for conjoint data
Can anyone tell me how to go about calculating the statistical reliability of conjoint data?
Thanks
Dave Hussey
*****************************************************
David Hussey
MORI
Tel: 0171 928 5955
Internet Address: [log in to unmask]
Home Page: http://www.mori.com
*****************************************************
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|