|
|
Hi subscribers,
I am using the PROPOR plugin that was recently added to the SPSS Development Site but cannot find documentation as to what kind of confidence intervals PROPOR calculates. It is stated that it calculates binomial (and poison) CI’s but I cannot determine exactly what kind of confidence limits are calculated. When testing a few scenarios I found that the (supposedly binomial) intervals that match PROPOR the best (but not completely) are based on Bayesian statistics. For sure it does not seem to be Wilson Score, Wald, normal approximation, or Exact Clopper Pearson.
I wonder if anyone could shed light on this one?
Kind regards,
JMollerup
|
Locked
