Software packages and textbook methods can often give us a single 'number' that tells us how good a potential model is (or not) for the data we get from observing a process. But ... you need to understand that this is by no means a 'perfect' way of trying to understand your process. Relying on software to think for you means that you don't understand what is going on. And that means your analysis won't be based on an understanding of what is going on either.