My hypothesis is whether or not music with violent overtones, such as "gangsta rap", increases violent behavior among young males.

Which of the methods used for analyzing survey data would you use and why? Correalational or Regressional statistical analysis

I would use neither. I would use two groups, experimental (listening to rap) and control (listening to non-rap) with a Z or t-test.