These are the common methods of thesis argumentation. Relatively speaking, data is used more, because data is the most likely to cause * * * and the most convincing, so some good articles basically have experimental data and special laboratory matching.