That's not true. Part of science is doing experimentation. If an experiment doesn't reproduce, you need to find out why that is. The hypothesize and test part of the scientific process is every bit as much science as the rest, even if your tests show that an idea is wrong, or that more is going on then you thought.