multivariate goodness of fit test
- From: "deep_sea" <deakar@xxxxxxxxx>
- Date: 30 Aug 2006 02:38:15 -0700
Hi everyone,
I need help on multivariate goodness-of-fit. I have a data set
consisting of 143 observations from 48 dimensional vectors. and i want
to test whether the data gives me enough information to assess that the
underlying population is the one i estimated. I want to add that my
observations are discrete, so the distributions to be tested are
discrete. I should estimate the parameters from the data, but besides,
i don't want to specify the parameters in the null hypothesis. Is the
usual chi-square test suitable for multivariate data?
Thanks in advance,
deep_sea
.
- Prev by Date: creating interaction term in multilevel analysis using difference scores
- Next by Date: Correlation ratio = eta-squared, or eta?
- Previous by thread: creating interaction term in multilevel analysis using difference scores
- Next by thread: Correlation ratio = eta-squared, or eta?
- Index(es):
Relevant Pages
|
