Historically, the conduct of science has been broadly divided into theoretical and experimental work. In the 20th century, information technology created what may be regarded as a third, middle way of working.
Doing theoretical work in physics or chemistry once meant working with pencil on paper or with chalk on a blackboard. While that image still has validity today, a huge part of the work is now done with powerful computers. Today, scientists have access to staggering computational power that their predecessors in the days of Einstein and Bohr could not have imagined.
The credibility of any theory is based on how consistent its predictions are with experiments conducted in the real world. Unfortunately, many experiments are extremely difficult and/or expensive to conduct. This Student Science article does a good job of explaining how computers, running programs written by scientists, can help to test theories and to guide traditional experimental work.