How does a Correlational experiment differ from a quasi experiment?

I think it has to do with the quasi you cannot randomly assign people to groups and cannot infer causality. With correlational you are simply examine the relationship between two nominal variables.