Dear All,
Does anyone know what sample size it is best to use for both
regressions and cluster analysis:
- if we perform logistic regression, is it better to use a 50/50 sample
of respondents/non respondents and what sample size should we use? let'
s assume we have a mailing universe of 100,000 with a response rate of
7%
- if we have a customer base of 100,000 customers and want to perform
cluster analysis (hierarchical / non hierarchical), what sample sizes
should we use for both.
The above analysis would be done using sas or spss.
A big thank you in advance for your help.
Regards
--
|