The explanatory power of Ecosim simulated model

[shuran wang]

Dear all,

       I have constructed a historical model using Ecosim.  AIC and SS were used to assess the goodness of fit. But these methods are all relative standards. For example, I've got 4 models with SS of 5000 to 6000. The number seems very large, and it doesn't tell me wether my model is good enough.

       I've got two questions needs some answer:

1. Is there an absolute criteria for a good fitted model? Like how could I know whether my model is persuasive with enough explanatory power?
   
2. Does anybody know how to calculate the amount of variation the model explained for the observed data?

       Thank you very much

Best regards,

Cindy

[Upendra Bhele]

I think you have to reduce the  SS because it to high. Best model should fit to the observed data. With the Vulnerabilities and group info you can decrease the SS value. may be this can help you. 

All the best