There are plenty of statistical texts (and even Wikipedia and web sites) that explain hypothesis testing in the context of fitting a model. That seems like the place for you to start reading. I'll try to explain briefly, but really, you can't learn statistics in the SAS Community, you need to take classes and read and study.
The null hypothesis is almost always that there is no effect (or in laymen's terms, the model does not find any real structure, the data is all noise).
The p-value of <0.0001 indicates a very low likelihood that this result happened by random chance under the null hypothesis (or in layman's terms, something that is not noise is detected and modeled).
And is it positive or negative? Is what positive or negative? Actually, positive or negative doesn't even apply here.
--
Paige Miller