Compare statistically significant difference between model output & observations

jones123

New member
Joined
Jun 7, 2013
Messages
6
Hi all,

I have created a model that simulates the occurrence a specific type of vegetation at 3 different places 'KU', 'KO' and 'N'. Now I want to compare the model results with observed data. I got the following table (attachment).

Which statistical test can I use to determine whether there is a significant difference between observed and modeled data?

Thanks!

I tried the Chi-squared test with H0: there is no significant difference. Chi squared was 104 so that H0 was rejected at a = 0.05. But this can't be true if you look at the data?

obsmod.png
 
Top