AI-GEOSTATS: t-test for spatial data
- Hi, all:
I would appreciate your comments on following approach:
Suppose that I have n data points s(l), l =1,..n.
If I like to test whether the mean is zero, traditionally I would take the
sample mean, sbar, and sample variance, say sigma1sbar, to use the t-test,
such as t=sbar/sigma1sbar.
However, if the data is observed from 2-d space and the data are spatially
correlated, then this t-test would inflate the type-I error. What I propose
to do is to use the variogram model to find the correlation among data
points and then the sigma2sbar would be then calculated from the
In this case, the sigma2sbar would be always bigger than the sigma1sbar
since it incorporates the spatial correlation and the t-test would tend to
accept the null hypothesis.
Any comments and suggestions for this approach? Any references?
* To post a message to the list, send it to ai-geostats@...
* As a general service to the users, please remember to post a summary of any useful responses to your questions.
* To unsubscribe, send an email to majordomo@... with no subject and "unsubscribe ai-geostats" followed by "end" on the next line in the message body. DO NOT SEND Subscribe/Unsubscribe requests to the list
* Support to the list is provided at http://www.ai-geostats.org