[ferret_users] statistically significant correlation

[Steeven Paul Yerraguntla <steevenpaul@xxxxxxxxx>]

Dear ferreters,
        I have plotted contours of correlation coefficients using "go variance".  I wanted to plot contours of statistically significant correlation coefficients (r).. For that i used a thread in ferret archives  http://www.pmel.noaa.gov/maillists/tmap/ferret_users/fu_2009/msg00482.html.  To assign degrees of freedom (f), i need to mention no of points (N). In case of one dimensional variables i can directly give N value. How to assign no of points(N) in case of Spatial correlation plot. 

         In student's t-test, as for as i know, t values should not exceed critical t value, ie., ttest values should be <= (le) to ctricital t (from table). I am sorry, if i am wrong, please correct me.  But in the quoted weblink there it was mentioned that it is statistically significant if  ttest >=(ge) critical t.

Suppose i got variable sign (significant used in thread), will i have to multiply it with correl variable in order to get variable spatially significant correlation?
Please advice me how can i clear my doubts and plot contours of statistically significant correlaion in ferret.   Thanks in advance.

regards,
- Paul.
       

--

Steeven Paul Y                              
Research Fellow
National Institute of Oceanography(NIO)
Regional Center, Visakhapatnam (RCV)                 
176 Lawsons Bay Colony
Visakhapatnam - 530017