Well, when a scientist begins working on a problem they need information so they aren't "going in blind," I guess you could say. The information they gain before working on problem (which usually involves an experiment or two) helps them to gain a better understanding of what they need to do to find an adequate solution.
