I focus on qualitative research with actual users to create simple, solid solutions.
Interactive design is about behaviour. I think interactive problems can only be correctly and with more certainty be indentified, tested and solved by observing real users performing realistic tasks in a (close to) natural setting using actual content.
Even low-cost user research through interviews and observation is highly effective for coming up with useful and intuitive solutions. I prefer to stay away from "blind" statistics, as they often provoke bias, unsupported opinions and assumptions.
I like to use a strong theoretical basis to come up with smooth, practical solutions. Especially after my studies I've bought and read books on interaction design, product and game design, including: