I've always viewed the act of computer programming as a bunch of experiments or thought experiments. You first build a hypothesis in code, then you evaluate it and analyze the results.

I like this because it puts the science in computer science.