Here’s an incredibly economical depiction of the relationship between evidence, experiment and theory swiped from Three Quarks Daily. It shows both the process and the uncertainty. It’s a little on the long side, so I’m placing it below the fold.
One aspect that this doesn’t cover is the usefulness of a theory. Theories are used to make predictions that can be tested. Those tests are themselves experiments.
To follow the analogy, we’d use the pentagon shape figure out where a point that we haven’t yet discovered might lie. We look there and try to find a point. If it’s there, our confidence in the theory is strengthened. If it isn’t, we go back several steps, fiddle with the drawing and try again. (As Kuhn pointed out, it isn’t quite that simple, but close enough for this example.)
Isaac Asimov once used the analogy of constructing a map using points of latitude and longitude. The advantage of this analogy is that you can immediately see how to make useful predictions with a map. Figure our where you want to go and try driving there. If you wind up in the middle of the ocean, you know your map is inaccurate. Make corrections and try again.
The question arises: is the map the terrain? Or in other words, is the theory a perfectly accurate depiction of the reality? The pure answer is: of course not. The pragmatic answer is: who cares?! As long as the map gets you to where you want to go, it’s good enough.