Can you run repeatable and observable experiments to explain phenomena? Can you combine present and past experimental results in larger theories? Yes, biology is a science. Still hate it though, living things are messy and disgusting...
Anyway, the XKCD comic was probably way over her head. The purity asserted in it was something else entirely, something akin to the reverse order of discovery, and on what other fields a branch of science is based upon. For example, us chemists are nothing more than scientists who apply the principles of physics to elements, atoms, molecules, polymers, etc, while the physicists find applications to equations solved centuries ago by mathematicians. A much better example would be quantum mechanics, which wasn't "discovered" until the early 1900s. Only that wasn't true in the slightest. The necessary equations had been solved starting in the 1700s, long before anything was known of the wave/particle nature of tiny things. Again, physics merely found an application for these equations, which us chemists then used in large scale computations for various properties of compounds, which the bio people then get to use for whatever the hell it is they're doing.
That said, sociology is just glorified statistics, and psychology and voodoo are interchangeable terms. (just bitter cuz the liberal arts people get so much money)